MONETARY POLICY AND ECONOMIC PERFORMANCE – EVIDENCE FROM

SELECTED AFRICAN COUNTRIES

n
w
To
Vera Ogeh Lassey Fiador
e
ap
C
of
ity

Thesis Presented for the Degree of

rs

DOCTOR OF PHILOSOPHY
ve

in the Graduate School of Business

ni

UNIVERSITY OF CAPE TOWN

U

PROMOTER:
PROFESSOR NICHOLAS BIEKPE

December, 2015
 n
The copyright of this thesis vests in the author. No

w
quotation from it or information derived from it is to be

To
published without full acknowledgement of the source.
e
The thesis is to be used for private study or non-
ap
commercial research purposes only.
C
of

Published by the University of Cape Town (UCT) in terms

of the non-exclusive license granted to UCT by the author.
ity
rs
ve
ni
U
 DECLARATION
I, VERA OGEH FIADOR, hereby declare that the work contained in this thesis is my own
original piece of work and that I have not previously, in its entirety or in part, submitted it at any
university for any degree.

Signature removed

Signature… Date: December, 2015

VERA OGEH LASSEY FIADOR

i
 ABSTRACT
The main aim of this dissertation is to broaden the understanding of the monetary policy
transmission mechanism as it operates in Sub-Saharan Africa. The ultimate goal is to aid in the
appropriate design and implementation of monetary policy for the attainment of developmental
goals. The dissertation empirically explores four issues on the pricing, behavioural and output
implications of monetary policy. The four questions that the dissertation attempts to answer in
Chapters Two to Five respectively are: 1) Does the effectiveness of monetary policy transmission
depend on the financial development of an economy? 2) Can monetary policy be used as a tool
in easing pressure on domestic currencies in the foreign exchange market? 3) Do banks engage in
excessively risky behaviour when monetary policy is expansionary? 4) Does monetary policy
influence aggregate variables like private capital formation, growth and their interrelationships?

Chapter Two tests the completeness of the pass-through of the central bank policy rate to bank
lending rates on one hand, and the interest rate pass-through as a function of the level of
financial development on the other hand. The results show that the pass-through of central bank
policy rate to bank lending rates is asymmetric for three Anglophone West African countries,
namely: Gambia, Ghana and Nigeria, which are seeking to ascend onto a single monetary
framework. However, there is no evidence that financial development affects the pass-through
of monetary policy. These findings still prove relevant, especially with regard to the quest for
effective monetary policy implementation and the ascension onto a single monetary framework
by these 3 countries. The motivation for this study stemmed from policy discussions and
academic debates on the premise that financial development is a key element in the pursuit of
effective monetary policy implementation, focusing on the three Anglophone West African
countries between 1975 and 2011. The study employs the bounds testing approach to
cointegration, and the Autoregressive Distributed Lags (ARDL) by Pesaran et al., (2001). The
findings show significant differences in the interest rate pass-through of the 3 countries studied.
Ghana and Gambia were characterised by undershooting in the response of lending rates to
monetary policy changes whilst Nigeria was characterised by overshooting in bank lending rates.
Financial development proved significant in some, but not in all the cases, while economic
growth proved mostly insignificant in the transmission of the policy rate to bank lending rates

In Chapter Three, we show that contractionary monetary policy of high interest rates is able to
correct disequilibrium in the foreign currency market in selected countries in Sub-Saharan Africa
(SSA). The chapter also provides empirical evidence about the impact of macroeconomic

ii
fundamentals on the domestic foreign exchange market. The study assesses the impact of
monetary policy on foreign exchange market pressure (EMP) in developing country contexts
focusing on some selected countries in SSA. EMP is the sum of exchange rate depreciation and
change in foreign reserves that is required to restore equilibrium to the domestic foreign
exchange market. The study was motivated by the fact that most of the SSA countries are
developing economies that have negative net export positions and stand to lose significantly
from consistently deteriorating foreign exchange positions. This study thus sought to measure
the ability of monetary policy to significantly address currency pressures that arise from trading
on the global market. The hypothesis that a tighter monetary policy stance can lend strength to a
currency was tested in this study using Generalised Methods of Moments (GMM) estimation in a
dynamic panel setting. Data on 20 SSA economies for which data were available for the period
1991 to 2010 are used. The study found a negative and significant relationship between
monetary policy and EMP, implying that contractionary monetary policy can ease EMP. It also
revealed significant relations between aggregate output, levels of public debt, the current account
balance and terms of trade and EMP. The findings of the study prove relevant as regards the
policy direction on exchange rate and currency management.

In Chapter Four, the study fails to confirm the presence of the risk-taking channel of monetary
policy in SSA in the light of the hypothesis that expansionary monetary policy can trigger risky
behaviour by banks. Contrary to this hypothesis, the findings seem to indicate that banks in SSA
were more likely to engage in risky behaviour when monetary policy is contractionary, especially
in an environment of rising interest rates. The study was motivated by the financial crises of
2007 – 2009, which seemed to point to excessive risk-taking by banks in periods of low interest
rates and ultimately the possibility of financial and economic instability. The chapter tests the
hypothesis that expansionary monetary policy increases the risk-appetite and risk-tolerance of
banking institutions in Sub-Saharan Africa (SSA) and causes an increase in risk-taking behaviour.
The study employed a dynamic panel model estimated using the Arellano and Bond difference
Generalized Method of Moments (GMM) estimator with bank-level data on 91 banks in 12 SSA
countries, spanning 1999 – 2012. Both macro-level and bank-level characteristics also proved
statistically significant in determining bank risk-taking behaviour. On the whole, the findings
present intricate policy implications, given SSA’s need for significant capital accumulation via
low interest rates to help stimulate economic growth.

iii
In the last chapter, Chapter Five, we show that monetary policy and private capital accumulation
are relevant drivers of growth. We, however, fail to establish a statistically significant relationship
between monetary policy rate and private capital accumulation for the selected sample countries
in SSA. The study employed data on 9 countries in SSA for the period 1999 – 2011 to test the
relationship between the central bank monetary policy rate, economic output and private capital
formation for Sub-Saharan Africa. The approach was the system Generalized Method of
Moments (GMM) estimation in a dynamic panel setting. The results of the study provide
empirical evidence to support the traditional view of a long-run positive relationship between
capital accumulation and economic growth. The findings also reveal an uncharacteristic positive
effect of contractionary monetary policy on growth, contradicting the conventional argument
that contractionary monetary policy hinders growth. Private capital formation, on the other
hand, is unaffected by monetary policy changes while the stock market proves insignificant in
terms of its influence on private capital formation and economic growth. The share of foreign
trade in GDP shows ambiguity: being positively significant for private capital but negatively
significant for growth. The findings of this chapter hold complex policy implications for SSA.

Overall, this thesis argues that monetary policy is relevant in the pursuit of growth and
development. The effectiveness in design and implementation, however, are conditioned by the
presence of other facilitating macro and micro factors.

iv
 DEDICATION
Giving all glory to God, I dedicate this thesis to my husband, Daniel and our three girls:
Lorraine, Norelle and Clare-Shan.

v
 AKNOWLEDGEMENTS

Thanks be to God Almighty, a phase in my life that seemed like it would never end, finally has.
My heartfelt appreciation goes to my supervisor Professor Nicholas Biekpe. Having come to
know you, I have found both a mentor and a father. Your comments have always had a way of
bringing out the best in me. I am and will be forever grateful to you for taking me on as your
student.

To my loving husband, Daniel Lassey Fiador, I owe you a world of gratitude. You believed in me
and always gave me the push to fly higher. I have said it time and again, if I had to marry all over
again, I would still choose you. I also acknowledge my senior colleagues, especially Josh and
Charles – your feedback and constant tabs on my progress were invaluable. May the good Lord
replenish all you expended on my behalf. To Dr. Simon K. Harvey, you gave real meaning to the
saying ‘You do not know what you can do until you are pushed’. Thank you for pushing me,
God bless you. Mrs. Eme Umoeka Fiawoyife, unknowingly, you set me on the path of academia,
thank you and God richly bless you.

Special mention to my girls for allowing me to take away their ‘mummy-time’ ever so often; to
my parents for laying the foundation blocks; to Auntie Caroline Kitcher, for setting the pace; to
Godwin Niikoi Kotey (may his soul rest in peace) for lighting the PhD flame in the family but
not living to finish his or see the fire blazing; and to my siblings for filling in for me whenever I
had to be away. To Willie Insaidoo and Emmanuel Abbey, thank you for demystifying the
econometrics.

I also wish to acknowledge the University of Ghana Business School and the Ghana Education
Trust Fund for financing my studies.

In fact, so many were the helping hands that I cannot mention them all, but to all who
contributed to this PhD reality, I pray that the good Lord bless you now and always. Thank you
all.

vi
 TABLE OF CONTENTS
DECLARATION .............................................................................................................................. i
ABSTRACT ..................................................................................................................................... ii
DEDICATION ................................................................................................................................ v
TABLE OF CONTENTS ............................................................................................................. vii
LIST OF FIGURES ........................................................................................................................ ix
CHAPTER ONE ................................................................................................................................1
INTRODUCTION .............................................................................................................................1

1.1 Background to the Study ................................................................................................................................. 1

1.2 Problem Statement ........................................................................................................................................... 2
1.3 Research Objectives ......................................................................................................................................... 4
1.4 Significance of the Study.................................................................................................................................. 4
1.5 Limitations of the Study .................................................................................................................................. 5
1.6 Structure of the Dissertation ........................................................................................................................... 5
CHAPTER TWO............................................................................................................................... 7
MONETARY POLICY AND FINANCIAL DEVELOPMENT IN ANGLOPHONE WEST AFRICA
.......................................................................................................................................................... 7

2.1 Introduction ....................................................................................................................................................... 7

2.2 Literature Review .............................................................................................................................................. 9
2.2.1 Monetary Policy Effectiveness ................................................................................................................ 9
2.2.2 Financial Development and Macroeconomic Pursuits ..................................................................... 10
2.2.3 Monetary Policy Effectiveness and Financial Development ........................................................... 11
2.3 Methodology and Data .................................................................................................................................. 13
2.3.1 Hypothesis and Empirical Model ......................................................................................................... 13
2.3.2 Model Estimation.................................................................................................................................... 16
2.3.2.1 Main Variables of Interest .................................................................................................................. 17
2.3.2.2 Control Variables ................................................................................................................................. 17
2.3.3 Data ........................................................................................................................................................... 17
2.4 Findings and Discussions .............................................................................................................................. 18
2.4.1 Some Stylized Facts: Financial and Economic Structure of Sample............................................... 18
2.4.2 Country Level Analysis........................................................................................................................... 22
2.4.2.1 Model Estimations and Discussion by Country ............................................................................. 22
Gambia .................................................................................................................................................................... 23
Ghana ...................................................................................................................................................................... 28
Nigeria ..................................................................................................................................................................... 32
2.5 Conclusion and Policy Implications ............................................................................................................ 35

vii
CHAPTER THREE......................................................................................................................... 38
MONETARY POLICY AND EXCHANGE MARKET PRESSURE - EVIDENCE FROM SUB-
SAHARAN AFRICA ....................................................................................................................... 38

3.1 Introduction ..................................................................................................................................................... 38

3.2 Literature Review ............................................................................................................................................ 40
3.2.1 The Monetary Policy Transmission Mechanisms ................................................................................... 40
3.2.2 Monetary Policy Transmission in Emerging Market Economies (EMEs) ......................................... 40
3.2.3 The Exchange Rate Channel of Monetary Transmission ..................................................................... 41
3.2.4 Determinants of Exchange Market Pressure .......................................................................................... 42
3.3 Methodology and Data .................................................................................................................................. 46
3.3.1 Sample and Source of Data ................................................................................................................... 46
3.3.2 Model Motivation.................................................................................................................................... 47
3.3.4 Estimation Methodology ....................................................................................................................... 48
3.4 Results and Discussions ................................................................................................................................. 49
3.4.1 Correlation Analysis ................................................................................................................................ 49
3.4.2 Regression Results................................................................................................................................... 51
3.5 Conclusion and Policy Implications ............................................................................................................ 55
CHAPTER FOUR ......................................................................................................................... 57
MONETARY POLICY AND BANK RISK-TAKING BEHAVIOUR IN SUB-SAHARAN
AFRICA - SOME EMPIRICAL EVIDENCE .............................................................................. 57

4.1 Introduction ..................................................................................................................................................... 57

4.2 Literature Review ............................................................................................................................................ 59
4.2.1 The Risk-Taking Channel of Monetary Policy ................................................................................... 59
4.2.2 Determinants of Bank Risk-Taking Behaviour ....................................................................................... 61
4.4 Findings and Discussions .............................................................................................................................. 70
4.4.1 Correlation Analysis ................................................................................................................................ 70
4.4.2 Regression Results................................................................................................................................... 72
4.5 Conclusion and Policy Implications ............................................................................................................ 79
CHAPTER FIVE ............................................................................................................................ 81
MONETARY POLICY, PRIVATE CAPITAL FORMATION AND ECONOMIC GROWTH –
THE CASE OF SUB-SAHARAN AFRICA ................................................................................... 81

5.1 Introduction ..................................................................................................................................................... 81

5.2 Review of the Literature ................................................................................................................................ 83
5.2.1 Interest Rates and Private Investment ................................................................................................. 83
5.2.2 Private Investment and Economic Performance ............................................................................... 85
5.2.3 Interest Rates and Economic Output .................................................................................................. 86

viii
 5.3 Methodology .................................................................................................................................................... 87
5.3.1 Modelling and Estimation Approach................................................................................................... 87
5.3.2 Justification of Variables ........................................................................................................................ 89
5.3.2.1 Central Bank Policy Rate .................................................................................................................... 89
5.3.2.2 Stock Market Development ............................................................................................................... 89
5.3.2.3 Private Capital Formation................................................................................................................... 90
5.3.2.4 Trade ...................................................................................................................................................... 90
5.4 Findings and Discussion ................................................................................................................................ 90
5.4.1 Descriptives and Correlation Analysis ................................................................................................. 91
5.4.2 Regression Analysis................................................................................................................................. 92
5.5 Summary, Conclusion and Policy Implications ......................................................................................... 98
CHAPTER SIX ............................................................................................................................. 100
SUMMARY, CONCLUSION AND RECOMMENDATIONS .................................................. 100

6.1 Introduction ................................................................................................................................................... 100

6.2 Summary of Purpose and Findings ............................................................................................................ 100
6.3 Recommendations ........................................................................................................................................ 102
6.4 Conclusions and Recommendations for Future Research ..................................................................... 104
Appendix A .................................................................................................................................... 134
Appendix B: Regressions on sensitivity to model specifications .......................................................... 136

LIST OF FIGURES

Figure 2.1: Trends in GDP for Gambia, Ghana and Nigeria............................................................. 18

Figure 2.2: Trends in financial development for Gambia, Ghana and Nigeria ............................... 19
Figure 2.3: Trends in central bank policy rate for Gambia, Ghana and Nigeria ............................. 20
Figure 2.4: Trends in bank lending rates for Ghana, Gambia and Nigeria ...................................... 21

LIST OF TABLES
Table 2.1 Summary Statistics of Interest rate volatility over study period per country .................. 22
Table2.2: Results of unit root tests ......................................................................................................... 22
Table 2.3: ARDL output on financial development and monetary policy: Gambia ....................... 24
Table 2.4: ARDL output interacting financial development and monetary policy: Gambia ......... 27
Table 2.5: ARDL output on financial development and monetary policy: Ghana ......................... 29
Table 2.6: ARDL output interacting financial development and monetary policy: Ghana ........... 30
Table 2.7: ARDL output on financial development and monetary policy: Nigeria ........................ 32
Table 2.8: ARDL output interacting financial development and monetary policy: Nigeria .......... 34

ix
Table 3.1: Correlation matrix of study variables .................................................................................. 50
Table 3.2: Regression output including South Africa .......................................................................... 52
Table 3.3: Regression output excluding South Africa ......................................................................... 54
Table 4.1: Correlation matrix of study variables .................................................................................. 71
Table 4.2: Regression results for main model: Bank risk as a function of micro and macro
characteristics............................................................................................................................................. 73
Table 4.3: GMM-IV regression output with interaction of bank liquidity and policy rate ............ 76
Table 4.4: Regression output with interaction of bank capital and policy rate ................................ 78
Table 5.1: Summary Statistics of Study Variables (1999-2011) .......................................................... 91
Table 5.2: Correlation Matrix of Study Variables ................................................................................. 92
Table 5.3: Regression output for growth in SSA .................................................................................. 94
Table 5.4: Regression Results for capital formation for SSA ............................................................. 95

x
 CHAPTER ONE

INTRODUCTION

1.1 Background to the Study

Endogenous growth theory argues that economic growth is facilitated by fixed capital
accumulation (Barro, 1991). Fixed capital accumulation, which is believed to stimulate
productivity, is also facilitated by a developed and efficient financial system. A financial system
is well-developed if it is resilient and able to withstand asset price fluctuations that result from
substantial increases in uncertainty and demand and supply conditions. The finance-growth
nexus, which is the main thrust of this study, is premised on the ideology that well-developed
financial systems are able to churn out various products to meet the demands of the productive
sector. In other words, a developed financial system is one that facilitates productivity and
ultimately increases aggregate output. This presupposes that an efficient financial system that
facilitates capital accumulation is likely to promote economic growth. Such a clear consensus is,
however, empirically lacking and so the debate about the role of financial factors in the pursuit
of economic development still rages on

In managing financial systems, the key policy tool is monetary policy. This suggests that the
efficiency and stability of the financial system, together with capital accumulation, are outcomes
of the monetary policy stance and its effects within an economy. For instance, on the issue of the
volume of credit that can be accessed by the productive sector of the economy or supplied by
the financial markets, a key determinant is the cost of the funds. The cost of funds is also usually
a function of the monetary policy stance within the economy. Mishkin (2007) argues about the
importance of monetary policy for the financial stability and health of any economy. A
discussion about an efficient and developed financial system that can promote growth will
therefore be incomplete without investigating the role of monetary policy. While this conceptual
link between monetary policy, capital accumulation and economic growth seems obvious, the
transmission dynamics between financial factors such as monetary policy and the real economy is
still plagued with a lot of empirical and ideological issues.

Over the years, many developing countries, Sub-Saharan African countries included, have been
faced with a myriad of problems in their pursuit of sustainable levels of growth and
development. Most of these problems have revolved around price instability, deteriorating

1
currencies, deteriorating current account positions with its attendant decreases in foreign
reserves, just to mention a few. In the 1980s and 1990s, the governments in Sub-Saharan Africa
embarked on various structural reforms aimed at revitalizing the economies, with financial sector
reforms at the core. The financial sector reforms across developing countries were possibly
informed by the arguments that successful economic development was typically preceded or was
accompanied by a major episode of financial innovation.

The general policy instrument adopted by most of these countries to combat the shocks to the
economy was monetary policy. Some of the prescriptions for monetary policy included lending
strength to currencies on the exchange market; stimulating capital formation and growth;
generating additional foreign reserves or both, while staving off financial instability or
contracting the money supply to slow down demand-related inflation. On one hand, it is well
noted that contractionary monetary policy can result in undesirable outcomes which are clear
opposites to the initial expectations, such as currency depreciation, high unemployment rates and
stagnation of the general economy. On the other hand, a very expansionary monetary policy of
very low interest rates with very low inflation levels that fails to stimulate aggregate demand
poses the risk of financial instability because it makes cash holdings more attractive in
comparison with interest-bearing bank deposits (see Stein, 1998; Chiesa, 2001; van den Heuvel,
2002; Herrero and Lopez, 2003; Diamond and Rajan 2006).

There is a consensus, according to Gamber and Hakes (2005), that monetary policy matters for
the economy. While this is the case in developed economies (Angeloni et al., 2003; Starr, 2005;
Bhattacharya et al., 2011; to mention but a few), the same cannot necessarily be said for
developing countries mainly due to a dearth of literature for developing economies and also
because of structural and institutional differences between developing and developed economies
that constrain the applicability of a developed country findings for developing countries. Indeed,
Bernanke and Gertler (1995) indicate that the transmission mechanism of monetary policy still
remains ‘a black box’ because the exact dynamics of the transmission of monetary policy to the
economy still remain largely unknown.

1.2 Problem Statement

As indicated earlier, in the 1980s, a number of the countries in Sub-Saharan Africa embarked on
a process of major economic reforms, typically referred to as the Structural Adjustment
Programme (SAP). The motive for the reforms was to increase and sustain economic growth in

2
these countries through price and macroeconomic stability. The main features of the reforms
included cuts in government spending, removal of import controls, privatization, currency
devaluations and tight-fisted control of money supply. Of the structural reforms pursued, the
financial sector reforms were key, mainly motivated by the financial repression model formulated
by McKinnon (1973) and Shaw (1973). The financial sector reforms were core to the structural
reforms also because, according to the World Bank Institute (2013), by the end of the 1980s,
many low-income countries, SSA included, faced unsustainable amounts of both domestic and
foreign debt. This caused investment stagnation, choked off economic growth, dropped social
spending and increased the suffering of the populace. In the quest to raise long-term capital to
finance both government and business activity and stimulate economic growth, these economies
pursued the financial reforms. These reforms notwithstanding, Sub-Saharan Africa has failed to
converge, in terms of growth, with the developed world (Mwega, 2003). This situation has given
rise to many, often diverse and conflicting, explanations on the genesis and nature of the
developmental difficulties in Africa – ranging from structuralist schools of thought to
neoclassical monetarist analyses.

The transmission mechanism of monetary policy typically evolves in stages. According to

Pétursson (2001), the first stage of monetary transmission is the financial system channel where
monetary policy actions transmit to interest rates, asset prices and exchange rate of the domestic
currency. The next stage encompasses the transmission from the financial system to the rest of
the economy, i.e. the effect on spending decisions, aggregate demand and ultimately inflation.
Studies exploring the various stages of the transmission abound in the literature. Of the studies
on the impact of monetary policy on interest rates (Sander and Kleimeier, 2004; De Bondt, 2005;
Sørensen and Werner, 2006); on asset prices (Filardo, 2000; Bernanke and Gertler, 2000;
Gilchrist and Leahy, 2002; Rigobon and Sack, 2004; Geromichalos et al., 2007); on the exchange
rate setup (Connolly and Da Silveira, 1979; Tanner, 2001; Gochoco-Bautista and Bautista, 2005;
Gali, and Monacelli, 2005; Iacoviello, 2005); on spending decisions (Chowdhury et al., 1986;
Gaiotti and Generale, 2001; McCarthy and Peach, 2002; Erceg and Levin, 2006); on aggregate
demand (Chari et al.,1995; Christiano et al., 2005; Bhattacharya et al., 2009; Uhlig, 2005) and on
inflation (Rabanal, 2007; Gupta et al., 2010), however, not many have focused on Africa and even
less on Sub-Saharan Africa, in particular. This gap in the literature poses a number of problems;
notable among these is the lack of appropriate design of monetary policy for sustainable growth
and development.

3
In view of this gap in the literature on SSA and the fact that the core of the numerous economic
reforms in Sub-Saharan Africa have been financial in nature and centred on monetary policy
dynamics, the empirical questions that come to mind are, ‘is monetary policy design and its
implementation part of the solution to Africa’s slow growth problems?’ Does monetary policy
really offer the answers for rapid economic growth through its impact on the financial system
and capital accumulation, for example? The study is also motivated by the fact that Sub-Saharan
Africa is yet to attain the desired level of economic development. This study seeks to provide
answers to some of the questions raised and in the process add to the extant literature on the
transmission mechanisms of monetary policy in the context of developing countries in SSA. The
study also seeks to shed some light on the question of the appropriateness of monetary policy as
a tool in the pursuit of enhanced economic performance. The findings of this empirical work
hold important implications for the appropriate policy responses for developing countries, for
issues like pressure on currencies, for bank behaviour and financial stability, for capital
accumulation and economic performance on the whole. Here economic performance is a broad
metric that encompasses issues of economic growth, financial sector development and stability,
effectiveness of policy implementation and capital accumulation.

1.3 Research Objectives

The main objective of the study is to explore the finance-growth nexus in SSA with the focus on
monetary policy transmission mechanism. The specific objectives of the study are:
 To examine the relationship between monetary policy transmission and financial
development in some selected countries in Anglophone West Africa.
 To determine the link between monetary policy and exchange market pressure in Sub-
Saharan Africa.
 To examine the relationship between monetary policy and bank risk taking behaviour in
some selected countries in Sub-Saharan Africa.
 To determine the relationship between monetary policy, private investment and
economic growth in Sub-Saharan Africa.

1.4 Significance of the Study

This dissertation examines the role of monetary policy in the economic performance of selected
countries in Sub-Saharan Africa. This it does by employing both macro and micro level data. As
one of the contributions to the literature, the study examines the ongoing debate on monetary

4
integration and the role of financial architecture in explaining the effectiveness of monetary
policy transmission. This is done by examining the differences in the interest-rate-pass-through
mechanisms of selected countries in Anglophone West Africa. A second contribution is that it
examines the interrelations between monetary policy and exchange market pressure. Yet another
contribution from this study is its findings on how macroeconomic fundamentals influence
domestic currencies in the selected countries in SSA. The study also contributes to the
discussions on how economic agents respond to monetary policy changes. This is achieved by
examining the role of monetary policy changes in explaining the risk-taking behaviour in the
banking sector. Last but not least, the study examines the tail end of the pass-through
mechanisms – assessing the relationship between monetary policy and economic output together
with intermediate targets, like aggregate investment, in the private sector.

Practically, it also seeks to add to the understanding of the monetary transmission mechanism so
as to facilitate the appropriate design and implementation of monetary policy for sustainable
growth and development in SSA.

1.5 Limitations of the Study

The focus of this dissertation is Sub-Saharan Africa. Despite this broad focus, the general lack of
datasets for all the countries within this sub region constituted the main limitation for the study.
On account of this data limitation, the richness of a large cross-country data could not be fully
explored. It also limited the span in terms of time, thus limiting some of the dynamics that could
have been gained from individual country analyses of some of the phenomena being studied.
Another limitation, also due to insufficiency of data, was the inability to cluster similar countries
so as to elicit more nuances in the response of the target phenomena to monetary policy
changes.

This data limitation, notwithstanding, the study presents some interesting findings that add to the
literature and stimulate further exploration into the ‘black box’ of the monetary transmission
mechanisms in Sub-Saharan Africa.

1.6 Structure of the Dissertation

Chapter Two presents the first essay of the dissertation, investigating the level of financial
development and the effectiveness of the interest rate pass-through. The third chapter presents

5
the second essay on monetary policy and exchange market pressure. The third empirical essay is
on the risk-taking behaviour of banks. Given expansionary monetary policy is presented in
Chapter Four. The fifth chapter investigates the relationship between monetary policy, capital
accumulation and growth. The last chapter, Chapter Six, presents the conclusions, implications
and suggestions for future research.

6
 CHAPTER TWO

MONETARY POLICY AND FINANCIAL DEVELOPMENT IN ANGLOPHONE

WEST AFRICA

2.1 Introduction

In the last two decades or so, the definition and measurement of the concept of financial
development has attracted a significant portion of the academic and policy literature. This
research interest has been kindled mostly by the notion that ‘finance leads growth’- a hypothesis
which seems to be winning the argument with most economists, despite the fact that the debate
on the finance-growth nexus still rages on. Quite a number of the studies that have been
devoted to financial development have identified its level as a key ingredient in the pursuit of
economic agendas. Bhattacharya and Sivasubramanian (2003), for instance, investigated the
possibility of causal relationship running from financial development to economic growth using
data spanning the period 1970 – 1999 for India and concluded that financial sector development
causes GDP growth. Chakraborty (2008) also found a long-run relationship between financial
development indicators (bank credit, stock market capitalization) and economic growth while
Katircioglu et al., (2008) also found support for a long-run relationship between financial
development and real GDP (see also King and Levine, 1993; Khan, 2001; Minella, 2001; Masten
et al., 2008, among many others).

Financial development has been generally defined as the policies, factors and institutions that
facilitate efficient intermediation and effective financial markets. According to Aghion et al.,
(2006), financial development is beneficial to economic growth at sufficiently high levels. This
benefit, they argue, works through the capacity of highly developed financial sectors to shield
economies from the negative effects of volatility that may stem from unfavourable shocks to the
economy.

In linking the value of financial development to the theory on convergence, Aghion et al., (2006)
further argue that any country that attains a certain critical level of financial development will
ultimately converge to the growth rate of the world technology frontier. Given the obviously
important role that the level of financial development seems to play in the quest for growth, can
it be hypothesized that the effectiveness of economic policies in stimulating an economy towards

7
it desired growth target is also influenced by the level of financial development? Is the
effectiveness of economic policy, say monetary policy, an attribute of the level of financial
development in an economy? Effectiveness of monetary policy is captured as the fullness and
speediness of pass-through of monetary policy to target variables like inflation, aggregate
demand, and market interest rates.

In answer to these questions and related ones on macroeconomic performance, a number of

studies have sought to measure the effectiveness of monetary policy transmission. Some have
argued that there is a link between the effectiveness of monetary policy in an economy and the
level of financial development that has been attained by the economy in question. In fact, a
number of such studies have concluded in one way or the other that the pass-through of
monetary policy is dependent on the level of financial development: the structure, the level of
competition, stages of financial development, among other characteristics (see Cottarelli and
Kourelis, 1994; Krause and Rioja, 2006, among others).

While these studies linking monetary policy effectiveness to the level of financial development
do not necessarily abound, they seem even more limited for a region like Sub-Saharan Africa; a
region that is in dire need of answers to its growth issues, ineffective policies and institutions.
This study is therefore one such attempt at filling this research gap, by empirically documenting,
for some selected countries in the Anglophone West African sub region, the relationship
between the level of financial development and monetary policy effectiveness, given the
implications for savings, investment and growth.

Studying the Anglophone West African Sub region is particularly relevant at this time. The
region presents a very opportune setting because of its quest for a unified monetary union. In
this regard, the assertion by De Bondt (2000) that the viability of a monetary union is partly
dependent on member states’ responsiveness to monetary policy makes such a study even more
pertinent and timely. This study seeks to highlight possible and relevant policy issues that may
need to be addressed in the quest for a successful single monetary policy framework. The study
contributes to the literature in the following ways: the first is to document the levels, differences
and the pace of financial development for the region over the past 3 decades. It also seeks to
examine the relationship between the level of financial development and monetary policy
effectiveness. The findings of this study thus seek to provide answers to relevant questions as

8
regards the effectiveness of monetary policy for the region and how financial development may
be harnessed in the pursuit of policy effectiveness.

As argued by Gigineishvili (2011), knowing what factors drive market responses and bank
behaviour to central bank actions would have valuable implications for strengthening monetary
policy effectiveness. This knowledge could also provide important input for the choice of a
monetary framework, including intermediate targets and policy instruments, and help identify
measures that are needed to improve it.

2.2 Literature Review

2.2.1 Monetary Policy Effectiveness

The concept of monetary policy effectiveness has gained a lot of research prominence in recent
times as economies around the globe continue to search for ways to improve the effectiveness of
economic policy in general. The importance of monetary policy effectiveness has increased
specifically on account of the need to determine whether central banks, of both developing and
developed countries, can exert strong and systematic effects on aggregate demand through
monetary policy. Policy, in this setting, is considered effective when it is capable of influencing a
predetermined target in the manner and intensity intended.

Testing for the effectiveness of monetary policy, despite having been around for a while, has
mostly focused on identifying which of the traditional transmission mechanisms of monetary
policy is prominent in a particular economic setting without much attention to the catalytic
factors. While monetary policy effectiveness seems visible as regards developed economies
(Mishra, Montiel and Spilimbergo, 2011), the same cannot be necessarily said of developing
economies. For the developed economies, for instance, the interest rate channel of monetary
transmission is perceived as the most effective, whereas the bank-lending channel is believed to
work for developing economies (Mishra et al., 2011). The bank-lending channel seems to be
considered the route for effective monetary policy in developing countries because these
economies are mostly bank dependent, while lacking the other frameworks, namely bond
markets, stock markets et cetera on which the other transmission mechanisms are deemed to
work.
According to Mishra and Montiel (2013), though a significantly large literature has emerged in
recent years with a focus on the empirical measurements of the effects of monetary policy on

9
aggregate demand and price levels, the beneficiaries have been mostly advanced economies. They
go on to intimate that most of these studies have tended to confirm the effectiveness of
monetary policy in influencing aggregate demand and prices, but are quick to add that there are
strong reasons to believe that similar effects may not necessarily hold for countries with
fundamentally different financial structures. This general conclusion has kindled research interest
into the determinants of monetary policy effectiveness as opposed to just the identification of
transmission mechanisms. Of the determinants in literature, levels of financial development have
received significant attention.

2.2.2 Financial Development and Macroeconomic Pursuits

The view that a well-structured financial system is important in boosting an economy has, for
some decades now, directed a significant amount of research towards the definition and
measurement of financial development. This has also fostered research aimed at identifying the
impact that financial development may have on macroeconomic variables. Studies linking the
financial sector and its level of development to various aspects of economic performance are
abundant and still adding up (King and Levine, 1993; Levine, 1997; Khan and Senhadji, 2000,
among many).

Financial sector development, according to DFID (2004), encompasses issues of

competitiveness and efficiency within the financial sector; the range of services provided; the
diversity of institutions within the sector; the volume of credit intermediated, along with access
to the services and financial stability. These characteristics, in turn, reduce transaction costs,
facilitate capital accumulation and ultimately increase aggregate output. Given the numerous
characteristics that come with financial sector development, its measurement, despite being of
significant research interest, has over the years defied a clear consensus. While some measures
are skewed towards developments in the banking sector, some are also skewed towards
securities’ markets with yet another set trying to deduce indices by combining characteristics of
all the parts that make up the financial sector.

Notwithstanding the apparent lack of consensus on how to appropriately measure financial

development, quite a number of studies have provided empirical evidence to support the notion
that financial development, irrespective of its mode of measurement, is important to the growth
and development debate. One of such early studies on financial development and growth is that
of King and Levine (1993). They find evidence that point to a strong and positive relationship

10
between the various indicators of financial development and growth, with the values of the
coefficient estimates implying that the impact is large. Levine, Loayza and Beck (2000) also
confirm that financial development exerts a large and positive influence on economic growth.
Susanto et al., (2011), using a sample of developed and developing countries documented a
positive relationship between the level of financial development and international trade, notably
exports. This positive relationship between financial development and exports is further
corroborated by Kiendrebeogo (2012), focusing on manufacturing exports, with evidence from
75 countries over the period 1971 – 2010. Berthelemy and Varoudakis (1996) also suggest that
when there is insufficient financial development, a country may find itself in a ‘poverty trap’.
This, they argue, can happen even if the economy has attained other conditions necessary for
sustained economic development. They find that countries with a high level of educational
attainment, but a low level of financial development, are saddled with relatively low standards of
living compared to other countries with similar levels of educational attainment but more
developed financial sectors.

From the foregoing, it seems quite obvious that financial development is important for
economic growth and development. Suffice it to argue that another area worth exploring, in
connection with financial development, is the monetary policy transmission mechanism.

2.2.3 Monetary Policy Effectiveness and Financial Development

The effectiveness of monetary policy as a research agenda has been around for some decades
now. The focus of such studies, however, has been on identifying the most effective monetary
transmission mechanism at work in (any) particular economy(ies). Over the years, it has emerged
from most of these studies that traditional monetary transmission mechanisms such as the credit
channel and the interest rate channel operate through the financial system (Carranza et al., 2010).
This suggests that the financial system is crucial to the effectiveness of the monetary
transmission mechanism. Carranza et al., (2010) go on to indicate that, this notwithstanding, the
theoretical and empirical literature has largely ignored the relationship between financial
development and monetary policy effectiveness. This is even so in spite of the wide variations
that have existed in the structures of financial systems across countries for decades now.

The proposition to determine the effects of the financial system on the monetary transmission
mechanism interestingly, seems to have been raised much earlier by Allen and Gale (1999) and
Cottarelli and Kourelis (1994). Allen and Gale (1999) believed that insights into the financial

11
structure of economies were of crucial importance because it had large potential effects on
economic efficiency in general and monetary policy transmission specifically. Cottarelli and
Kourelis (1994) furthered this argument by also arguing that the effectiveness of monetary policy
hinged on a “set of structural parameters not directly controlled by central banks”, one of which
was obviously the level of financial development. De Bondt (2000) later also buttressed this need
to look into the relationship between financial development and monetary policy effectiveness
by intimating that the imperfection of financial markets gives an essential role to the financial
structure in the monetary transmission mechanism. Financial development vis-à-vis monetary
transmission has only recently received the needed attention as a result of the need to identify
means of enhancing the effectiveness of monetary policy.

The empirical literature, which has been far and in between, has mostly concluded that the level
of financial development is crucial to the effectiveness of monetary policy. One of the
pioneering works in this area was by Cottarelli and Kourelis (1994). They concluded that
differences in interest rate pass-through in the Euro zone can be explained by differences in their
financial structures: banking system competition, extent of developments in the money markets,
ownership structure of banking institutions and barriers to foreign competition (see also Mojon,
2000). Cecchetti (1999) corroborated the significance of financial development in monetary
policy effectiveness by also reporting that “countries with many small and less healthy banks
with poorer access to direct capital displayed greater sensitivity to monetary policy changes than
did countries with big healthy banks and deep well-developed capital markets”. In another study
of 21 countries by Lastrapes and McMillin (2004), they found that where the financial markets
were developed and allowed economic agents to easily rebalance their portfolios in reaction to
monetary policy shocks, the liquidity effect of monetary policy was subsequently weaker. In a
relatively larger study of 37 industrialised and developing economies by Krause and Rioja (2006),
the finding was that a more developed financial market significantly contributed to a more
effective monetary policy implementation, irrespective of whether the economy was a developed
or a developing one. This finding is a contradiction to Cecchetti (1999) and Lastrapes and
McMillin’s (2004) negative relationship between financial development and monetary policy
effectiveness. Sorensen and Werner (2006) also found that competition within the banking
sector, a characteristic of developed financial markets, was positively significant in explaining a
speedier pass-through for the policy interest rate. In a recent study by Carranza et al., (2010), it
also emerged that while monetary policy has a larger cumulative impact when financial systems

12
are less developed, there are significant time lags between implementation and impact, while for
developed economies the experience was that of a speedier pass-through for monetary policy.

In general, the scant empirical literature definitely lends itself to the idea that financial
development is key to determining the effectiveness of monetary policy. The impact, however,
could be facilitative or dampening. In one dimension, a well-developed financial system provides
the structures for the transmission of policy to the economy, thus contributing to effectiveness.
On another dimension, well-developed financial markets could also provide insulation against
monetary policy shocks by providing numerous media/innovations by which economic agents
can counter the effects flowing from monetary policy shocks. This ambiguity in direction and the
fact that most of these studies are limited to the advanced economies still leave open the
question, how does financial development affect the effectiveness of monetary policy
transmission, and in this particular case, for Anglophone West Africa?

2.3 Methodology and Data

This section details the approach to the study. It touches on the empirical model, data types and
sources as well as the model estimation.

2.3.1 Hypothesis and Empirical Model

The basic generalization from the literature presupposes that the efficacy of monetary policy as a
stabilization tool is influenced in a large part by the strength and reliability of the linkage between
instruments of monetary policy and the target economic variable. The strength and reliability of
this linkage, according to Opolot et al., (2013), is also in part dependent on the country’s financial
architecture: the size, the composition of the formal financial sector, the degree of development
of money and capital markets, the level and degree of financial innovations and financial
integration and the external payments regime. Based on the foregoing, this study seeks to test the
hypothesis that ‘the effectiveness of monetary policy is dependent on the financial development characteristics
within an economy’.
In testing for the efficacy of monetary policy in an economy, many studies have relied on the
strength and speed of the traditional transmission mechanisms (Gigineishvili, 2011). The
common feature of all the transmission channels is that they transmit policy decisions to
aggregate demand and/or supply through financial markets. In order to test for the pass-through
of monetary policy for the countries under study, the cost-of-funds approach, developed by de Bondt

13
(2005) is employed. The cost-of-funds approach is employed following John and Pokhariyal’s
(2013) argument that the stage of the pass-through from policy rate to bank retail rates is best
described using the cost of funds approach because the lending rate is the price of financial
goods provided by the banking sector. In following with Mojon’s (2000) study on the varying
impact of financial structure on the strength of the interest rate channel, this study models the
interest rate pass-through as also being dependent on financial development. In controlling for
other determinants in the pass-through studies (see Cottarelli and Kourelis, 1994; Sorensen and
Werner, 2004b, Sander and Kleimeier 2004 and Gigineishvili, 2011) the most common variables
have been inflation, volatility, economic growth and competition. Inflation, however, could not
be modelled mainly because of its high correlation with one of the important variables of
interest, (the policy rate, which could not be dropped). Similarly, competition could not be
modelled because of inconsistency of data series for the countries being studied. Consequently,
the interest rate pass-through is specified as a function of the central bank policy rate, financial
development, economic growth and interest rate volatility, giving the empirical model for the
study as:

LRt   0  1MPRt   2 BSIZEt   3GDP   4 BVOLt   t ………………(1)

In order to test if changes in monetary policy have differential impacts given varying levels of
financial development, equation (1) is re-estimated with an interaction term, MPR * BSIZE
following the intuition of the general literature on modelling the moderating effects of financial
development on pass-through (Cottarelli and Kourelis, 1994; Cecchetti, 1999; Mojon, 2000; de
Bondt, 2005; Krause and Rioja, 2006, and others). The expectation is that if financial
development influences the pass-through of monetary policy, then in equation 2, 3 ҂ 0. In

essence, the pass-through effect is 1 + 3 . As a robustness check, the financial development

variable is still included as an explanatory variable. This ensures that any no-zero effect from the
interaction term does not flow from an inappropriately specified equation. The augmented
model is thus given as:

LRt   0  1 MPRt   2 BSIZE t   3 MPRt * BSIZEt   4 GDP   5 BVOLt   t ………..….(2)

For both equations, LR is the average annual bank lending rate in each economy, MPR is the
central bank discount/prime rate to proxy for the monetary policy stance; BSIZE is the size of

14
the banking sector, proxied by the private credit to GDP ratio and the indicator of financial
development within an economy while GDP and BVOL measure the overall health of the
macroeconomic environment and banking industry volatility respectively. MPR*BSIZE (in
model 2) is the interaction term between financial development and monetary policy.

Following from the cost of funds approach, β1 (the coefficient for the monetary policy) is
expected to follow the path of 0≤ β1≥1. A value close to zero implies that the pass-through is
ineffective and sluggish while a value equal to one (1) is indicative of the completeness and
effectiveness of monetary policy. In instances where β1 is greater than one (1), the situation is
described as over-shooting and according to de Bondt (2005) is indicative of banks’ adjustment
to control for the decrease in probability of repayment in periods of contractionary monetary
policy.

While the conventional argument has a positive sign as the a priori expectation for financial
development (BSIZE), the empirical pass-through literature has exhibited more of mixed signs.
On the one hand, because high levels of financial development provide the appropriate
structures for the transmission of policy moves, financial development is expected to reflect
positively on pass-through. On the other hand, where high levels of financial development
provide innovative platforms to circumvent the implications for central bank rate changes, the
sign is expected to be negative. Samba and Yan (2010) show that when financial market
developments are characterised by imperfections, like lack of banking sector competition, excess
banking sector liquidity and a poor structure (stock markets), it serves to impede pass-through.
This seems to suggest that unless developments in the financial sector promote competition its
contribution to pass-through is compromised. On the basis of these, the financial development
variable is expected to exhibit either a positive or negative relationship with lending rates.

According to Gigineishvili (2011), how banks respond to interest rate movements is largely
determined by the degree to which they believe that interest rates carry reliable information
about underlying market trends. Interest rate volatility, however, introduces uncertainty into
market signals, causing banks to interpret interest rate movements with more caution.
Consequently, rather than immediately passing the changes in market rates to customers, they
choose to wait until the noise is filtered out. If this happens, interest rate volatility is expected to
exhibit a negative effect with lending rates. This path notwithstanding, high levels of volatility
will cause banks to adjust lending rate significantly upwards to compensate for the decreased

15
probability of repayment whilst maximising their risk-adjusted returns and protecting their
profitability. In this case, volatility will be expected to have a positive relationship with bank
lending rates. In view of the differing paths of volatility transmission to lending rates, there is no
a priori expectation in terms of the coefficient sign.

It is also argued that, in periods of rapid economic growth, ceteris paribus, there is an attendant rise
in the demand for credit. This ultimately creates competition in the financial sector. This
competition in response to the increased loan demand can compel banks to easily pass on
downward changes in the central bank policy rate to lending rates (Egert et al., 2006), where loan
demand is elastic. If the increase in loan demand becomes inelastic, however, then lending rates
are likely to be characterised by rigidities and sluggishness in response to policy rate changes. In
terms of empirical evidence, however, the path is not so clearly mapped out. Sander and
Kleimeier (2004), on the one hand, found economic growth to be insignificant in the short-run
but significant in the long run. Gigineishvili (2011), for instance, found per capita GDP to be
insignificant in explaining the interest rate pass-through mechanism. He interpreted the
insignificance as possibly indicative of the cyclicality of GDP, which banks do not factor into
their pricing decisions. Based on the available and relatively inconclusive empirical evidence,
there is no a priori expectation with regard to the coefficient sign.

2.3.2 Model Estimation

The empirical model was run as a time-series analysis. This choice flows from the major
drawback that arises in relation to cross-country studies; they generate estimates of the average
effects, though the relationship of interests may vary considerably between countries. Fedderke
(2005) argues that “where general laws genuinely hold, then they must be evident not only in
aggregate, but in particular instances as well”. This presupposes that studies of particular or
individual cases become useful because they allow for circumstances and factors that may be
unique to the specific instance under study to be controlled for with a degree of precision that
may not be possible for aggregate cross-country studies. This intuition has stimulated a number
of studies examining the pass-through nexus employing country-level time series data and some
(for example Odedokun 1996; Demetriades and Hussein 1996 among others) have actually
found variations in the relationship that can be attributed to country-level dynamics. On account
of the countries’ desire to ascend onto a single monetary framework, the model is estimated with
an a priori expectation that the relationships being tested will exhibit synchronicity across the

16
sample countries, and serve as attestation to their readiness to mount the single monetary
framework.

2.3.2.1 Main Variables of Interest

Private credit to GDP is used as the proxy for measuring financial development for the study
sample. Financial development is measured this way primarily because the SSA region’s financial
markets are dominated by banks. The second reason for using private credit is because it is the
standard metric for measuring developments in the banking system. High levels of financial
development, all things being equal, are expected to enhance the effectiveness of monetary
policy. The expectation is that the pass-through of monetary policy to retail rates will be larger
and swifter. The central bank policy rate/prime rate following Delis and Kouretas (2011) is used
as the measure for monetary policy rate, . As per their study, the choice of interest rate
measure does not lend much difference to the outcome. The central bank rate is used also
because it is the de facto indicator of the monetary policy stance in the economy given and also
because it is the primary tool used by central banks for implementing monetary policy. The
monetary policy stance is expected to exhibit a direct relationship with bank retail rates.

2.3.2.2 Control Variables

In order to account for other factors, aside the variables of interest that may exert influences on
the monetary transmission mechanism, a set of controls are included: a measure of the
macroeconomic environment, proxied by GDP growth and market volatility, proxied by the
standard deviation on the policy rate. Apart from measuring the level of economic activity, GDP
is also quite informative with regard to other macroeconomic characteristics. GDP is expected to
enhance pass-through. The use of the standard deviation of the policy rate to proxy market
volatility follows Gigineishvili (2011). Market volatility is included as a control to assess the
effects of market uncertainty and the information flow dynamics in the financial market setting –
a significant factor in the pass-through of policy moves.

2.3.3 Data
Annual data covering three out of the six Anglophone West African economies (Ghana, Nigeria,
and The Gambia), for which data are available, are sampled for this study. These Anglophone
economies in West Africa are on the path of integrating their monetary frameworks and

17
therefore present a unique dataset for the study. The study period is from 1975 to 2011. Data on
all the variables are obtained from the World Bank’s Development Indicators (WDI) and the
IMF’s International Financial Statistics.

2.4 Findings and Discussions

This section, divided into two main parts, deals with the findings of the study. The first part
presents some stylized facts on the countries under study, specifically on the economic and
financial structure over the sample period. The second part presents the country-level analysis on
variable unit roots and the regression analysis.

2.4.1 Some Stylized Facts: Financial and Economic Structure of Sample

Economic growth for the 3 countries under study has waxed and waned over the 37-year period
from 1975-2011 (see Figure 2.1).

Figure 2.1: Trends in GDP for Gambia, Ghana and Nigeria

GDP Growth
20

15

10

GAMBIA
5
Percent

GHANA
0
NIGERIA
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011

-5

-10

-15

Economic Growth_GDP Mean Std. Dev. Min Max

Gambia 3.752162 3.528698 -4.3 12.39
Ghana 3.552703 4.837447 -12.43 14.39
Nigeria 3.462274 5.227168 -13.12788 10.6
Data compiled by author from WDI

18
Gambia recorded the highest average growth for the sample at, 3.752%, for the whole period
with its highest negative growth being -4.3% in 2011. Within the 37-year period, the country also
experienced annual growth of 12.39% in one period. Ghana followed closely behind with an
average growth of 3.552%, its highest negative growth being -12.43% and its maximum positive
growth being 14.39% over the period of study.
Nigeria had the least mean GDP growth of 3.462% whilst also recording the highest negative (-
13.128%) and lowest positive (10.6%) growth within the sample. In terms of highest GDP
growth in one fiscal year, Ghana took the lead with its record 14.39% in 2011.

Financial development (Figure 2.2), measured as the ratio to GDP of credit to the private sector,
also exhibits some divergence across the countries over the period under study. To start with, all
the countries had initially low levels of financial development. This improved markedly for
Gambia and Nigeria over the late 1970s, through to the middle and late 1980s.

Figure 2.2: Trends in financial development for Gambia, Ghana and Nigeria

Credit to the Private Sector (% of GDP)

45
40
35
30
Percent

25 GAMBIA
20 GHANA
15 NIGERIA
10
5
0
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011

Financial Mean Std. Dev. Min Max

Development_ BSIZE
Gambia 12.97703 6.568595 3.73 25.12
Ghana 7.651892 4.944318 1.54 15.88
Nigeria 14.72873 6.798227 6.814002 38.59011
Data compiled by author from WDI

19
From this point on, credit to the private sector, which also proxies for financial depth began to
deteriorate. Gambia, for example, has not been able to return to the level of financial depth that
characterized the era of the 1980s. Nigeria, on the other hand was able to pick up the pace,
peaking at 38.59% around 2009 before dropping once more to approximately 21% in 2011.

The story for Ghana, though, has been somewhat different. Prior to the IMF-led financial sector
liberalization programmes, the Ghanaian financial sector obviously lacked depth, evident from
the very low levels of private credit to GDP at the time. After the reforms in the late 1980s,
financial depth in Ghana has witnessed a consistent upward trend, implying increasing financial
development. Suffice to say that Ghana still recorded the lowest period average of 7.652% for
the measure of financial development and the highest period average was by Nigeria, 14.729%.

In terms of interest rates (Figure 2.3), the policy rate for Ghana showed the highest peaks over
the period. Gambia followed closely behind with Nigeria lagging behind. In terms of averages,
Ghana, Gambia and Nigeria recorded 21.96%, 14.964% and 11.716% respectively for the policy
rate.

Figure 2.3: Trends in central bank policy rate for Gambia, Ghana and Nigeria

Policy Rate
50

40
Percent

30 GAMBIA
20 GHANA
10 NIGERIA
0

Policy Rate_MPR Mean Std. Dev. Min Max

Gambia 14.96432 6.333497 6 29.25
Ghana 21 10.35105 8 45
Nigeria 11.71622 5.246084 3.5 26
Data compiled by author from WDI

20
In the last decade or so, however, Gambia recorded relatively high levels in its prime rate, in
comparison to the decade before, while the other two countries, Ghana and Nigeria experienced
relative declines in the rates over the same period. Between 1994 and 2002, the prime rate for
Ghana diverged significantly from those of Gambia and Nigeria, though the same period saw
more of a convergence between Nigeria and Gambia in their prime rates. From 2003 onwards,
Gambia experienced higher policy rates compared to Ghana and Nigeria.

The lending rate for Ghana, just like the policy rate, showed the highest peaks over the 37-year
period. The convergence between the Nigerian and Gambian policy rates was not so evident as
that of the lending rate. The lending rate for Nigeria was relatively lower than in the other two
countries over the sample period. In the year 2004, lending rates in all 3 countries began to
experience some decline. This decline continued for Gambia. The other 2 countries, however,
did not experience this consistency in the decline of their commercial bank lending rate.

Figure 2.4: Trends in bank lending rates for Ghana, Gambia and Nigeria

Bank Lending Rate

50
45
40
35
30
Percent

GAMBIA
25
20 GHANA
15 NIGERIA
10
5
0
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011

Lending Rate_LR Mean Std. Dev. Min Max

Gambia 24.18919 5.636379 14 37
Ghana 27.78378 8.310419 19 47
Nigeria 16.16177 6.389534 6 31.65
Data compiled by author from WDI

Volatility in the policy rate (see the Table 2.1) was also highest for Ghana, averaging 3.75 over
the period, followed by Gambia at 2.6 and Nigeria at 2.0. In general, Nigeria displayed the least
variability with regard to interest rates over the sample period.

21
Table 2.1 Summary Statistics of Interest rate volatility over study period per country

Interest rate volatility_BVOL Mean Std. Dev. Min Max

Gambia 2.628226 2.303575 0 8.871302
Ghana 3.75164 1.830891 .9797959 8.064738
Nigeria 2.057864 1.276957 0 5
Data compiled by author from WDI

2.4.2 Country Level Analysis

2.4.2.1 Model Estimations and Discussion by Country

In order to test the interest rate pass-through for the sample countries, their data series were first
subjected to the Augmented Dickey-Fuller and Phillips-Perron tests of unit root to ascertain
their order of integration, as this was key to the appropriateness of the autoregressive distributed
lag (ARDL) approach that was employed for the study. The ARDL approach, though not
requiring the series to be stationary in levels, is not applicable to series integrated of order 2 or
above. The unit root table below (Table 2.1) shows that the series for all 3 countries are
integrated of order 1 and below.

Table2.2: Results of unit root tests

ADF PP ADF PP
LEVEL 1st DIFFERENCE
NIGERIA
t-Statistic t-Statistic t-Statistic t-Statistic
DR -2.375193 -2.143998 -2.920130* -7.265499*** I(1)
LR -2.142630 -1.870137 -6.559456*** -6.569828*** I(1)
MPR -2.455623 -2.317772 -8.091393*** -8.410575*** I(1)
BSIZE -1.930660 -1.961486 -5.245551*** -6.671918*** I(1)
BVOL -2.776801* -2.649276* -7.469877*** -11.04667*** I(1)
GDP -5.12988*** -5.28999*** -10.89798*** -12.27508**** I(0)
GHANA
t-Statistic t-Statistic t-Statistic t-Statistic

22
 DR -2.494185 -2.410461 -7.049917*** -7.188982*** I(1)
LR -1.522339 -1.569858 -5.436549*** -5.440068*** I(1)
MPR -1.681625 -1.712045 -6.737009*** -6.689303*** I(1)
BSIZE -0.271095 -0.151687 -6.715834*** -6.723649*** I(1)
BVOL -1.644078 -1.931408 -6.285033*** -5.265712*** I(1)
GDP -3.91429*** -3.92425*** -6.595241*** -8.583349*** I(0)
GAMBIA
t-Statistic t-Statistic t-Statistic t-Statistic
DR -2.510405 -2.063369 -2.838321* -5.022258*** I(1)
LR -1.763751 -1.763751 -6.018453*** -6.018453*** I(1)
MPR -1.90905 -1.774799 -5.537147*** -3.986938*** I(1)
BSIZE -3.26010** -1.376169 -5.075021*** -5.075021*** I(1)
BVOL -1.018895 -1.946150 -3.876228** -4.639237*** I(1)
GDP -6.03006*** -6.73547*** -4.175552*** -10.82130*** I(0)
Note: ADF=Augment Dickey Fuller Test, PP=Phillip-Perron Test, I=Order of Integration, I(0)= Integrated at levels, I(1)
=Integrated at order 1, ***, ** and * denotes 1%, 5% and 10% level of significance for rejection of null hypothesis of unit root

After subjecting the series to unit root tests, the next step in the ARDL methodology was the
OLS estimation of models 1 and 2 in order to test for the presence of a long-run (level)
relationship. All three countries rejected the null hypothesis of no long-run relationship at 95%
following the F-statistics provided by Pesaran et al., (2001).

The regression models are thus estimated using the Autoregressive Distributed Lags with error
correction modelling. This approach is preferred over the other error-correction estimation
methods like the Engle-Granger two-step (1987) and Johansen (1988), mainly because it does
not require that the series be integrated of the same order and also because of its good small
sample properties (Bahmani-Oskoee and Hegerty, 2009). The ARDL model selection was based
on the Akaike information criterion, mainly because of its good small sample size properties. The
diagnostic tests (Tables 2.3-2.8) reveal no anomalies with the estimated models for all the three
countries. Thus, we proceed to present the findings.

Gambia
From Table 2.3, the coefficient of the monetary policy variable is positive and significant,
suggesting the presence of a pass-through effect. From Table 2.3, the pass-through, however, is

23
incomplete, standing at approximately 54%. This suggests that only 54% of an increase in the
policy rate is transmitted to bank lending rates. By inference, the policy rate must change by
twice the rate change that is expected in bank retail rates. This finding is in consonance with a
number of the studies that also document interest rate pass-through for developing countries as
being incomplete (Bhattacharya et al., 2011). They argue that this is because their financial
markets are also under-developed. In the short-run, as well, the coefficient is less that unity,
implying that changes in monetary policy are not completely transmitted to the lending rate.
With reference to both the short-run and long-run coefficients, it is apparent that the interest
pass-through is (quite) incomplete in Gambia.

The coefficient of the level of financial development, contrary to major findings in the literature
(see Krause and Rioja, 2006; Singh et al., 2008; among others), is negatively and significantly
signed in relation to the lending rate. This may suggest that increasing levels of financial
development impede interest rate pass-through for Gambia. This negative effect of financial
development on the interest rate pass-through could also be indicative of bottlenecks in the
financial system, specifically the banking system in Gambia hence the negative effect. On
account of the Cottarelli and Kourelis’ (1994) argument that competition and efficiency in the
financial sector are important for pass-through, the negative effect of financial development on
pass-through could be an indication that developments in the financial markets of Gambia are
not necessarily characterised by competition and efficiency, thus hampering the interest rate
pass-through. The negative sign and statistical significance persist even for short-run changes in
the level of financial development. In general, these findings could also be an indication that the
financial sector in Gambia is uncompetitive and possibly concentrated.
Contradictory to the literature as this finding may be, however, it seems to find support with the
finding by Carranza et al., (2010) that financial development can be negatively related to the
effectiveness of monetary policy.

Table 2.3: ARDL output on financial development and monetary policy: Gambia

Dependent F-statistic 95% 95% 90% 90% Conclusion

Variable D(LR) Lower Upper Lower Upper
Bound Bound Bound Bound
Model 1 10.25059*** 2.850 4.049 2.425 3.574 Reject at
95%

24
 Autoregressive Distributed Lag Estimates
ARDL(1,2,0,2,1) selected based on Akaike Information Criterion
Diagnostic Tests
Test Statistics LM Version F Version
Serial Correlation CHSQ(1) = .55840 [.455] F( 1, 23)= .37290 [.547]
Functional Form CHSQ(1) = 4.0443 [.044] F( 1, 23)= 3.0049 [.096]
Normality CHSQ(2) = 1.4169 [.492] Not applicable
Heteroscedasticity CHSQ(1) = .24174 [.623] F( 1, 33)= .22951 [.635]
Variable Long-run Coefficients Variable Short-run coefficients
MPR .53704 [.000]*** D(MPR) .33189 [.001]***
D(MPR(-1)) -.15063 [.159]
BSIZE -.32282 [.000]*** D(BSIZE) -.43763 [.000]***
BVOL .43753 [.070]* D(BVOL) -.19918 [.405]
GDP .089695 [.388] D(GDP) .12215 [.072]*
D(GDP(-1)) .11827 [.066]*
C 18.9907 [.000]*** D(C) 25.7446 [.000]***
ECM (-1) -1.3556 [.000]***
***,**,* indicate significance at 1%, 5% and 10% respectively. MPR is central bank policy rate and proxy for monetary policy;
BSIZE is private credit to GDP ratio and proxy for financial development; BVOL is interest rate volatility; GDP is real GDP
growth; ECM is error correction term. The model was subjected to the following diagnostic tests: the Breusch-Godfrey test for serial
correlation, the Ramsey’s RESET test for functional form, the Jacque-Bera test for normality and the White test for heteroscedasticity.
The diagnostic tests suggest that the above model has no significant autocorrelation, the functional form is appropriate, the errors are
normally distributed and the disturbances are homoscedastic.

Volatility of market rates, for Gambia, shows a significantly positive relationship with the lending
rate. From the perspective of uncertainty, it appears increasing volatility leads to increased
lending rates (increasing risk premium), possibly as a protective mechanism by banks against
potential declines in profitability. Though the sign is analogous to the expectations of the pass-
through literature (Sander and Kleimeier, 2004), another view from the market efficiency
perspective shows that volatility on the market actually filters into lending rate determination. In
the short-run, however, market volatility shows the expected negative sign but it is statistically
insignificant.

25
GDP shows the expected positive sign but it turns out to be statistically insignificant. This seems
to suggest that the interest rate pass-through is, in the long-run, divorced from macroeconomic
fundamentals like GDP. The statistical insignificance of GDP in the pass-through mechanism
could be flowing from threshold effects. In other words, there must be a certain level of growth
within an economy before the financial system becomes responsive to monetary policy action.
Short-run changes in GDP, as well as previous period changes, however, seem more relevant in
explaining pass-through in Gambia. The error correction term in this model is significantly
negative, implying a reversion to the long-run equilibrium lending rate after an exogenous shock
to the lending rate. The error-correction term has a larger than unity value, pointing to some
over-shooting in the reversion to long-run equilibrium lending rate.

Given that the speed of adjustment is equally informative on the magnitude of adjustment in
lending rates to central bank policy moves, the mean adjustment lag was computed, following
Hendry (2005). The mean adjustment lag for model 1 shows that for Gambia, bank-lending rates
take approximately 5.86 months (approximately half a year) to respond to central bank policy
rate changes. This indicates that policy moves must take into cognizance the half-year lag when
proposing and implementing policy directives. This adjustment lag obviously makes for a
sluggish pass-through.

In order to measure the effects of financial development on the pass-through mechanism, the
policy rate is interacted with the financial development variable (MPR*BSIZE) in model 2. The
long-run regression coefficients for model 2 for Gambia do not show a marked difference from
model 1 except for GDP that reverses its sign. Much consequence is not attached to this since it
remains statistically insignificant. The interaction term, MPR*BSIZE, exhibits a positively
significant relationship, suggesting that there is an enhancement of pass-through of policy in a
developed financial market. On the whole, however, it appears that financial development plays
only a marginally positive role in the pass-through for Gambia. The error-correction term
(ECM(-1)) is also statistically significant and negative, implying a reversion of lending rates to its
long-run equilibrium after an adverse shock. From the coefficient value, it appears that some
overshooting also occurs in this case.

26
Table 2.4: ARDL output interacting financial development and monetary policy: Gambia

Dependent F-statistic 95% 95% 90% Lower 90% Conclusion

Variable Lower Upper Bound Upper
D(LR) Bound Bound Bound
Model 2 10.34479*** 2.649 3.805 2.262 3.367 Reject at
95%
Autoregressive Distributed Lag Estimates
ARDL(1,0,1,1,2,2) selected based on Akaike Information Criterion
Diagnostic Tests
* Test Statistics *LM Version * F Version
Serial Correlation CHSQ (1) = 1.7002[.192] F (1, 21)= 1.0722[.312]
Functional Form CHSQ (1) = 2.5680[.109] F (1, 21)= 1.6628[.211]
Normality CHSQ (2) = 1.1707[.557] Not applicable
Heteroscedasticity CHSQ (1) = .071500[.789] F (1, 33)= .067553[.797]
Variable Long-run Coefficients Variable Short-run Coefficients
MPR .23416 [.069]* D(MPR) .32216 [.051]*
BSIZE -.49653 [.000]*** D(BSIZE) -.23824 [.227]
MPR*BSIZE .013249 [.067]* D(MPR*BSIZE) -.0076654 [.505]
BVOL .78792 [.004]*** D(BVOL) .020663 [.939]
D(BVOL(-1)) -.36727 [.185]
GDP -.014831 [.888] D(GDP) .029581 [.706]
D(GDP(-1)) .13238 [.048]*
C 22.8393 [.000]*** D(C) 31.4224 [.000]***
ECM(-1) -1.3758 [.000]***
***,**,* indicate significance at 1%, 5% and 10% respectively. MPR is central bank policy rate and proxy for monetary policy;
BSIZE is private credit to GDP ratio and proxy for financial development; BVOL is interest rate volatility; GDP is real GDP
growth; MPR*BSIZE is the interaction term for monetary policy and financial development; ECM is error correction term. The model
was subjected to the following diagnostic tests: the Breusch-Godfrey test for serial correlation, the Ramsey’s RESET test for functional
form, the Jacque-Bera test for normality and the White test for heteroscedasticity. The diagnostic tests suggest that the above model has no
significant autocorrelation, the functional form is appropriate, the errors are normally distributed and the disturbances are homoscedastic.

The mean adjustment lag for model 2, at 5.9 months (approximately 6 months), is also not
significantly different from that of model 1, confirming the fact that lending rates in Gambia take
approximately 6 months to adjust and reflect central policy changes in the monetary policy rate.

27
It would be expected that policy moves that are expected to reflect in bank retail rates must
either be made taking cognizance of the lagged effect or alternatively necessary step must be
taken to remove the bottlenecks in the banking sector that may impede the interest rate pass-
through

Ghana
The model 1 results for Ghana (Table 2.5) show a positively significant relationship between the
policy rate and the bank-lending rate. This coefficient value suggests that approximately only
48% of a unit increase in the policy rate filters into the bank-lending rate. In the short-run, the
pass-through to the lending rate is even more sluggish, evidenced by the fact that only 28% of a
change in the policy rate is immediately transmitted to the lending rate. This rigidity of interest
rate pass-through could also be an indication of a lack of credibility in the monetary policy
process, hence the significantly low pass-through.

The financial development variable, BSIZE exhibits a positive relationship with the lending rate.
By implication, developments in the banking sector enhance the pass-through of policy rates to
bank retail rate. It may be that in the long-run, developments in the banking sector promote
competition, facilitate symmetric information and discourage collusive behaviour, hence the
improvement in the pass-through. In the short-run, the financial development variable is
negative but not statistically significant. Market volatility for Ghana, like Gambia, also shows a
significantly positive relationship with lending rates. While contrary to the major findings in the
literature, it appears that lending rates in Ghana adjust upward in the presence of uncertainty. In
the short-term, however, volatility is inconsequential in affecting the pass-through.

The long-run and short-run coefficients for GDP growth were negative and insignificant. Thus,
like Gambia, economic fundamentals, like GDP growth, are divorced from the pass-through
process. Given that GDP is also a proxy for the overall economic conditions in a country, the
insignificance of GDP is a possible indication that bank-lending rates are influenced more by
factors external to the economy and less by internal factors. In essence, lending rates in Ghana
do not reflect macroeconomic growth conditions. This lack of statistical significance, according
to Gigineishvili (2011), could stem from the cyclicality of GDP which, he argues, banks do not
factor into their long-term pricing functions. GDP is also a proxy for aggregate demand so
another possible reason why GDP is not significant could be the operation of threshold effects.
Hence, without a certain minimum level of aggregate demand, lending rates are not responsive to

28
GDP. All in all, apart from the coefficient of short-run changes in monetary policy which was
significant, none of the other short-run variables were significant in model 1.

Table 2.5: ARDL output on financial development and monetary policy: Ghana

Dependent F-statistic 95% 95% 90% 90% Conclusion

Variable Lower Upper Lower Upper
D(LR) Bound Bound Bound Bound
Model 1 34.45689*** 2.850 4.049 2.425 3.574 Reject at
95%
Autoregressive Distributed Lag Estimates
ARDL(1,0,1,0,1) selected based on Akaike Information Criterion
Diagnostic Tests
* Test Statistics *LM Version * F Version
Serial Correlation CHSQ (1) = .0098756 [.921] F (1, 26)= .0073382[.932]
Functional Form CHSQ (1) = 4.3542 [.037] F (1, 26)= 3.6941 [.066]
Normality CHSQ (2) = 3.4122 [.182] Not applicable
Heteroscedasticity CHSQ (1) = 2.3218 [.128] F (1, 33)= 2.3447 [.135]
Variable Long-run Coefficients Variable Short-run Coefficients
MPR .48383 [.000]*** D(MPR) .28378 [.002]***
BSIZE .81989 [.001]*** D(BSIZE) -.30127 [.377]
BVOL 2.0019 [.017]** D(BVOL) .19197 [.704]
GDP -.12025 [.644] D(GDP) -.070532 [.643]
C 3.9592 [.150] D(C) 2.3222 [.196]
ECM(-1) -.58654 [.000]***
***,**,* indicate significance at 1%, 5% and 10% respectively. MPR is central bank policy rate and proxy for monetary policy;
BSIZE is private credit to GDP ratio and proxy for financial development; BVOL is interest rate volatility; GDP is real GDP
growth; ECM is error correction term. The model was subjected to the following diagnostic tests: the Breusch-Godfrey test for serial
correlation, the Ramsey’s RESET test for functional form, the Jacque-Bera test for normality and the White test for heteroscedasticity.
The diagnostic tests suggest that the above model has no significant autocorrelation, the functional form is appropriate, the errors are
normally distributed and the disturbances are homoscedastic.

The error-correction term for the model is highly significant, negative and less than unity. This
implies a stable long-run relationship between the lending rate and the policy rate, with a
reversion to equilibrium after an adverse shock. From the results of model 1 for Ghana,

29
approximately 59% of any anomaly in the lending rate that occurs in the prior period is corrected
in the current period. The speed of the pass-through for model 1 indicates that monetary policy
moves through changes in the policy rate by the central bank take approximately 14.65 months
(15 months) to filter through to lending rates. This overly slow speed of adjustment points to a
very ineffective interest rate pass-through in Ghana.

The results from model 2 with the interaction term MPR*BSIZE for Ghana (Table 2.6), show
that in a developed financial sector, both in the long-run and short-run, the transmission process
is enhanced. For the same model, the financial development variable is insignificant in the long-
run but significant and negative in the short-run. This could point to the fact that in the absence
of a financial sector that is capable of transmitting policy signals, policy rate changes are
ineffective in driving bank-lending rates and ultimately investment dynamics and growth. The
differences in effect of the single variable and the interacted variable could thus be interpreted as
indicating the need for alignment between policy actions and the characteristics of the economy.
This argument is proffered because it appears that in itself, financial development does not
unequivocally aid pass-through but when interacted it becomes significant both in the short-run
and the long-run.

Table 2.6: ARDL output interacting financial development and monetary policy: Ghana

Dependen F-statistic 95% 95% 90% 90% Conclusion

t Variable Lower Upper Lower Upper
D(LR) Bound Bound Bound Bound
Model 2 39.02940*** 2.649 3.805 2.262 3.367 Reject at
95%
Autoregressive Distributed Lag Estimates
ARDL(2,0,1,0,0,1) selected based on Akaike Information Criterion
Diagnostic Tests
* Test Statistics *LM Version * F Version
Serial Correlation CHSQ (1) = 1.8763[.171] F (1, 24)= 1.3595[.255]
Functional Form CHSQ (1) = .66918[.413] F (1, 24)= .46781[.501]
Normality CHSQ (2) = 6.1447[.046] Not applicable
Heteroscedasticity CHSQ (1) = .54121[.462] F (1, 33)= .51830[.477]

30
Variable Long-run Coefficients Variable Short-run Coefficients

D(LR(-1)) .20708 [.133]

MPR .0076760 [.962] D(MPR) .0062690 [.962]
BSIZE -.39886 [.359] D(BSIZE) -1.2321 [.019]**
MPR*BSIZE .070373 [.004]*** D(MPR*BSIZE) .057474 [.014]**
BVOL 1.2473 [.035]** D(BVOL) .18821 [.688]
GDP .090775 [.620] D(GDP ) .074136 [.624]
C 13.9932 [.000]*** D(C) 11.4283 [.005]**
ECM(-1) -.81670 [.000]***
***,**,* indicate significance at 1%, 5% and 10% respectively. MPR is central bank policy rate and proxy for monetary policy;
BSIZE is private credit to GDP ratio and proxy for financial development; BVOL is interest rate volatility; GDP is real GDP
growth; MPR*BSIZE is the interaction term for monetary policy and financial development; ECM is error correction term. The model
was subjected to the following diagnostic tests: the Breusch-Godfrey test for serial correlation, the Ramsey’s RESET test for functional
form, the Jacque-Bera test for normality and the White test for heteroscedasticity. The diagnostic tests suggest that the above model has no
significant autocorrelation, the functional form is appropriate, the errors are normally distributed and the disturbances are homoscedastic.

Volatility shows a positive relationship with lending rates in the long-run but is insignificant in
the short-run. This seems to suggest that short run variations in the policy rate do not filter into
bank lending rate determination but in the long run, such volatility seems to cause banks to
adjust rates upward to compensate for the uncertainty and noise level in the central bank policy
rate setting. From Table 2.6, it appears that the central bank rate has no statistically significant
relationship with bank lending rates. The insignificance could be reflective of asymmetric
information in the banking sector. On the other hand it could be reflective of collusive
behaviour on the part of banks since such collusive behaviour will result in a relatively long lag
before arriving at a new equilibrium price hence the lack of a significant relationship. Real GDP
remains insignificant.

The error-correction terms for both models are highly significant and negative and less than
unity, implying a stable long-run relationship with a reversion to equilibrium after an adverse
shock. In model 2, as much as 82% of any disequilibrium created in the previous period is
corrected in the current period. By inference from the results, the interest rate pass-through is
relatively ineffective, but improves in the presence of financial development.

31
Nigeria
Nigeria, contrary to its sister countries, is characterised by overshooting of the lending rate in
response to changes in the policy rate and this is significant at 1%. According to De Bondt
(2000), lending rates respond to changes in the policy rate by either overshooting or
undershooting. As per Sander and Kleimeier (2004) overshooting in lending rates can occur if
banks charge higher rates to offset the risks inherent in banking. Sorensen and Werner (2006)
also argue that overshooting may be an indication of significant information asymmetry between
lenders and borrowers.

Table 2.7: ARDL output on financial development and monetary policy: Nigeria

Dependent F-statistic 95% 95% 90% 90% Conclusion

Variable Lower Upper Lower Upper
D(LR) Bound Bound Bound Bound
Model 1 6.579069*** 2.850 4.049 2.425 3.574 Reject at 95%
Autoregressive Distributed Lag Estimates
ARDL(1,0,0,0,0) selected based on Akaike Information Criterion
Diagnostic Tests
* Test Statistics *LM Version * F Version
Serial Correlation CHSQ (1) = .65844[.417] F (1, 29)= .54029[.468]
Functional Form CHSQ (1) = 1.1310[.288] F (1, 29)= .94063[.340]
Normality CHSQ (2) = .93711[.626] Not applicable
Heteroscedasticity CHSQ (1) = 2.5838[.108] F (1, 34)= 2.6290[.114]
Variable Long-run Coefficients Variable Short-run
Coefficients
MPR 1.2203 [.000]*** D(MPR) .70939 [.000]***
BSIZE .20371 [.032]** D(BSIZE) .11842 [.034]**
BVOL .23285 [.724] D(BVOL) .13537 [.732]
GDP .10220 [.380] D(GDP) .059411 [.389]
C -1.8564 [.436] D(C) -1.0792 [.426]
ECM(-1) -.58134 [.000]***
***,**,* indicate significance at 1%, 5% and 10% respectively. MPR is central bank policy rate and proxy for monetary policy;
BSIZE is private credit to GDP ratio and proxy for financial development; BVOL is interest rate volatility; GDP is real GDP
growth; ECM is error correction term. The model was subjected to the following diagnostic tests: the Breusch-Godfrey test for serial

32
correlation, the Ramsey’s RESET test for functional form, the Jacque-Bera test for normality and the White test for heteroscedasticity.
The diagnostic tests suggest that the above model has no significant autocorrelation, the functional form is appropriate, the errors are
normally distributed and the disturbances are homoscedastic.

Inferring from the two opinions leads to the conclusion that the risks inherent in banking, as a
result of information asymmetry between borrowers and lenders, account for the overshooting
in the lending rates of the Nigerian banking sector. In the short run as well, there is evidence of a
pass-through at approximately 71% and this is significant at 1%. Financial development exhibits
a positive relationship with the lending rate, indicating that it facilitates the interest rate pass-
through both in the long and the short-run.

Market volatility and GDP growth are both insignificant, implying they are inconsequential in
explaining the pass-through in Nigeria. The insignificance of GDP is corroborated by the
findings of Egert et al., (2006). From the coefficient of the error correction term, approximately
58% of the disequilibrium in the lending rate in the previous period is offset in the current
period. The mean adjustment lag for model 1 for Nigeria is 5.99 months (approximately 6
months). This suggests that policy rates take about 6 months to be reflected in the cost of capital
of economic agents.

For model 2 (Table 2.8), the overshooting in the lending rate to changes in the policy rate
becomes even more pronounced, both in the short-run and long-run. In this instance, a unit
change in the policy rate reflects in more than a unit change in the lending rate. Upon interacting
the policy rate with financial development to measure the pass-through implication of monetary
policy as a function of developments in the financial sector, it appears, based on the significance
and the sign, that the pass-through is hampered in the presence of developments in the financial
markets. Both in the long-run and the short-run, the coefficients are negative and significant.
This could be an indication that increasing innovations in the financial sector of Nigeria provide
alternative means for market players to insulate themselves from monetary policy effects, thus
leading lending rates to exhibit a sluggish relationship to developments in the financial sector.

Another implication could be that the financial sector of Nigeria is underdeveloped, hence its
inability to transmit policy changes to lending rates, or that developments in the financial sector
are not necessarily characterised by competition and efficiency, the main drivers of the interest
rate pass-through. GDP growth and market volatility still remain insignificant. Gigineishvili

33
(2011) explains the insignificance of GDP growth in the pass-through framework as a possible
effect from GDP cyclicality which banks do not factor into their long-run pricing decisions.

The error term is significantly negative and less than unity, indicating a stable long-run
relationship that corrects approximately 60% of the previous period’s disequilibrium in the
current period. The speed of pass-through, inferred from the mean adjustment lag, is not
computed for this model as the short-run pass-though also exhibits overshooting.

Table 2.8: ARDL output interacting financial development and monetary policy: Nigeria

Dependent F-statistic 95%Lower 95% 90%Lower 90%Upper Conclusion

Variable Bound Upper Bound Bound
D(LR) Bound
Model 2 13.60962*** 2.649 3.805 2.262 3.367 Reject at
95%
Autoregressive Distributed Lag Estimates
ARDL(1,0,0,0,0,0) selected based on Akaike Information Criterion
Diagnostic Tests
Test Statistics LM Version F Version
Serial Correlation CHSQ (1) = .12368 F( 1, 28)= .096525 [.758]
[.725]
Functional Form CHSQ (1) = .018393 F( 1, 28)= .014313 [.906]
[.892]
Normality CHSQ (2) = .70360 Not applicable
[.703]
Heteroscedasticity CHSQ (1) = 1.6593 F( 1, 34)= 1.6429 [.209]
[.198]
Variable Long-run Variable Long-run
Coefficients Coefficients
MPR 2.2947 [.000]*** D(MPR) 1.3210 [.000]***
BSIZE .91519 [.001]*** D(BSIZE) .52688 [.000]***
MPR*BSIZE -.087808 [.004]*** D(MPR*BSIZE) -.050551 [.002]***
BVOL -.30658 [.628] D(BVOL) -.17650 [.614]
GDP .12853 [.205] D(GDP) .073997 [.214]

34
C -9.5183 [.008]** D(C) -5.4797 [.003]***
ECM(-1) -.59963 [.000]***
***,**,* indicate significance at 1%, 5% and 10% respectively. MPR is central bank policy rate and proxy for monetary policy;
BSIZE is private credit to GDP ratio and proxy for financial development; BVOL is interest rate volatility; GDP is real GDP
growth; MPR*BSIZE is the interaction term for monetary policy and financial development; ECM is error correction term. The model
was subjected to the following diagnostic tests: the Breusch-Godfrey test for serial correlation, the Ramsey’s RESET test for functional
form, the Jacque-Bera test for normality and the White test for heteroscedasticity. The diagnostic tests suggest that the above model has no
significant autocorrelation, the functional form is appropriate, the errors are normally distributed and the disturbances are homoscedastic.

Inferring from the formula for the mean adjustment lag, however, seems to indicate that lending
rates may actually pre-empt policy moves in Nigeria and thus adjust ahead of policy changes,
resulting in the over-shooting even in the short-run.

2.5 Conclusion and Policy Implications

The study sought to test the hypothesis that effective monetary policy transmission via bank
lending rates is a function of the level of financial development in an economy. Three countries,
Gambia, Ghana and Nigeria, within the West African Monetary Union, for which data was
available, were studied. Of the three countries studied, Nigeria exhibited the highest pass-
through, followed by Gambia and then Ghana. In the long-run, the depth of the financial sector,
as per evidence from the study, is somewhat important in explaining the transmission of interest
rates in the three countries, having exhibited differences in significance and mathematical signs.
Interestingly though, market characteristics like volatility and growth were not quite as
important, as far as the transmission mechanism was concerned. All three countries exhibited a
co-integrated relationship in the behaviour of lending rate. The long-run and the short-run
results from the error-correction modelling, using the ARDL, revealed significant divergences in
the pass-through responses of the three countries. Given that policy is considered effective when
it is capable of influencing a predetermined target in the manner and intensity intended, the
imperfect mapping of lending rates to policy rates is a clear indication of ineffective monetary
policy transmission to retail rates in the countries studied.
In terms of policy implications, the significant divergences among the three countries obviously
call into question the ascension on to a single monetary framework by these economies.
Following from De Bondt’s (2000) assertion that the viability of a single monetary framework
depends in part on the member states’ responsiveness to monetary policy, these disparities point
to a lack of readiness on the part of the economies to merge their monetary frameworks. There

35
is a need to re-evaluate the parameters and seek ways to address the underlying differences in
these member states.

Secondly, the evidence of incomplete pass-through for Gambia and Ghana brings to the fore
issues of monetary policy effectiveness. In essence, the inability of policy moves to transmit into
costs of capital for these countries implies that the durable sectors that are dependent on bank
financing cannot be manipulated via central policy moves. Put simply, there is the need to begin
exploring alternative monetary policy regimes that have the capacity to influence the cost of
capital and investment incentives of economic agents and ultimately economic growth. The
overshooting, on the other hand, that characterizes the Nigerian case also implies that in periods
of rising interest rates, costs of capital via lending rates will lead to a magnified reduction in
investments in the durable sector and ultimately shrink growth. On the one hand, periods of
falling interest rates, assuming lending rates are not sticky downwards, will create ease of access
to capital. This, however, brings with it the added risk of qualifying non-creditworthy firms due
to an increase in collateral valuations. This situation, in the event of an adverse shock to the
economy that triggers defaults, will ultimately predispose the banking sector to a financial crisis
and negatively affect the economy.

In the peculiar case of Ghana that is on an inflation-targeting regime of monetary policy, the lack
of responsiveness of bank lending rates to changes in the policy rate threatens the sustainability
of the policy regime. This is because a policy rate that is unable to induce changes in interest
rates so as to alter the effective cost of capital and the wealth of households and enterprises is, in
essence, unable to effect adjustments in consumption and investments and ultimately aggregate
demand and prices. It therefore calls for further probing into the efficacy of the prevailing
structural requirements that are expected to facilitate a policy regime like inflation targeting and
thus improve its efficacy. With regard to the effectiveness of pass-through in the presence of
financial development, it appears the findings of the study must be interpreted with caution
given the conflicting signs that characterised the results, and especially so because some of the
results pointed to negative pass-through implications of financial development. One insight,
though, is a perspective by Gigineishvili (2011) which indicates that an underdeveloped,
uncompetitive and inefficient financial market is actually an impediment to the pass-through
mechanism. This may be the case in Anglophone West Africa. In light of the findings, there is no
need to compromise on the quest to develop the financial sectors of the West African region,

36
but rather to take steps to improve the competition and efficiency of these markets so that they
serve the function of facilitating pass-through and ultimately economic growth.

37
 CHAPTER THREE

MONETARY POLICY AND EXCHANGE MARKET PRESSURE - EVIDENCE

FROM SUB-SAHARAN AFRICA

3.1 Introduction
In the wake of the globalization euphoria, the foreign exchange market has taken on a very
important role in the growth quest of many countries around the globe. Countries have had to
seriously manage their domestic currencies in relation to major trading currencies to attain and
sustain steady rates of long-term economic growth. This renewed attention is premised on the
argument that an appreciation or depreciation of domestic currencies plays a key role in
influencing a country’s trade balance (Stucka, 2004; Aziz, 2008) and ultimately its growth
(McPherson and Rakovski, 2000).

In light of this argument, countries have, over the decades, either sought to stimulate an
appreciation or depreciation in their domestic currencies in order to achieve some developmental
target. From conventional economic theory, currency depreciation makes a country’s exports
cheaper relative to foreign imports and is expected to promote economic growth by directing the
flow of capital into the domestic economy. This growth effect, however, is dependent on the
country having a positive net export position or, as Guitian (1976) and Dornbusch (1988)
illustrate, on the ability of the country to switch demand in the proper direction and magnitude.

Other theoretical arguments (see Cooper, 1971; Krugman and Taylor, 1978; and Barbone and
Rivera-Batiz, 1987) are, however, of the view that currency depreciation is more of a
contractionary, as opposed to being an expansionary move. From their perspective, consistent
currency depreciations may not necessarily lead to growth. It can, instead, lead to a currency
crisis when the exchange rate setup is poorly managed. A typical currency crisis results in loss of
trust in the ability of a country’s currency to act as a store of value. This directs the flow of
capital away from the domestic economy and negatively affects the currency. Excessive pressures
in the exchange market, according to Garcia and Malet (2007), can therefore be disastrous for a
country’s growth agenda.

Given the seemingly important role of a country’s currency to its growth objectives, a number of
studies have sought to unravel the determinants of exchange market pressure and ultimately

38
currency crisis. The motives have been for appropriate policy formulation. Parlaktuna (2005) in a
Turkish study spanning 1993 and 2004 reports that domestic credit is an important factor in
explaining the occurrence of exchange market pressure (EMP). Ucer et al., (1998) employed the
signals approach, also for Turkish data for the period from 1986 to 1999, and found that short-
term foreign debt and weak exports, relative to imports, increased the economy’s EMP and
ultimate vulnerability to currency crises. Studies by Girton and Roper (1977) and Connolly and
da Silveira (1979) on Canada and Brazil respectively suggest that growth in real GDP has an
appreciative effect on local currencies. Studying EU member states, Van Poeck et al., (2007)
found current account deficits and growth in domestic credit to be consistently significant in
explaining EMP. Hegerty (2010) later confirmed part of Van Poeck et al’s (2007) conclusions by
also reporting that a deterioration in current account deficits worsens the EMP. Kibritcioglu et
al., (1998) employed the leading indicators approach for the period from 1986 to 1998 for
Turkey and identified terms of trade as a leading indicator of a currency crisis. Tanner (2001) in a
cross-country study reported that contractionary monetary policy helps to alleviate currency
pressure. Gochoco-Bautista and Bautista (2005), however, are of the view that contractionary
monetary policy only alleviates currency pressure during tranquil periods, but during crisis
periods the effect is in the opposite direction. This suggests that contractionary monetary policy
deepens exchange market pressure when there is a prevailing currency crisis.

In general, the empirical results from these studies seem to point in the direction of
macroeconomic fundamentals as determinants of EMP. The empirics also strengthen Jayaraman
and Chee-Keong’s (2008) argument that, for developing countries, the determinants of EMP are
macroeconomic in nature. This argument notwithstanding, the studies on the determinants of
EMP are far and few between and where available are mostly skewed in the direction of
developed economies, making generalizations difficult. Furthermore, the available literature on
exchange market pressure (EMP) also offers no common ground with regard to sample
selection, methodology and variable choices. As a result, the findings vary along with the
different approaches in the literature. According to Feridun (2009), the heterogeneity of crises in
each country also makes the results of available studies unreliable for drawing specific
conclusions regarding the crises in other countries.

Probing the literature on exchange market pressure further reveals a dearth of literature to aid
policy design for Sub-Saharan Africa. This dearth in the literature is especially intriguing, as most
of the countries in Sub-Saharan Africa, in their quest to achieve and maintain sustainable levels

39
of economic performance have often alternated between the use of monetary or fiscal policy
variables. For this region, a question that has not been adequately addressed is whether the
monetary policy tools employed in the quest for growth and development have had the posited
effect on the target macroeconomic variables. Another question is whether these macro variables
exert some feedback that affects the conduct of policy.

This study is one such attempt to provide some empirical answers and help fill the gap for Sub-
Saharan Africa. The main objective of the study is to explore the monetary policy transmission
mechanism in some selected countries in SSA from the perspective of the foreign exchange
market. The study mainly seeks to test the hypothesis that a contractionary (expansionary)
monetary policy strengthens (weakens) a currency. The secondary objective is to identify which
macroeconomic fundamentals help to explain exchange market pressure in SSA.

3.2 Literature Review

3.2.1 The Monetary Policy Transmission Mechanisms

This section provides a brief review of the literature on monetary policy transmission
mechanisms in emerging economies and the determinants of exchange market pressure.
Monetary policy is generally defined as the measures and the set of procedures used by monetary
authorities to manage the supply of money, interest rates and exchange rates and to influence
credit conditions to achieve certain economic objectives. This area has witnessed significant
research interest in the past few decades. This interest, which is mainly focused on unravelling
the transmission mechanism of monetary policy, is, according to Meier and Muller (2005), due to
the fact that better insights into the nature of the transmission mechanism have obvious benefits
for policy-making. The literature in general seems to posit numerous mechanisms by which
monetary policy impacts an economy. The most traditionally acclaimed routes, however, are the
interest rate channel, the credit channel, the asset price channel and the exchange rate channel
(Mishkin 1996).

3.2.2 Monetary Policy Transmission in Emerging Market Economies (EMEs)

As per the literature, three distinct features of emerging economies make the monetary
transmission mechanism in such markets worth investigating. The first of such features is the
absence of a developed bond market which, according to Moreno (2008), makes transmission of
changes in the short-term policy rate to other points on the yield curve weak. The second feature

40
has to do with banking systems that are generally characterized by low levels of competition,
public ownership and high cost of information processing. These characteristics coupled with
‘the small size of the sector’, according to Bhattacharya et al., (2011) result in a small impact on
aggregate demand even if lending rates change in response to a policy move. The third feature
peculiar of emerging markets is a large informal finance sector. As a direct result of this feature, a
rise in central bank policy rate does not transmit well to the economy, because the shocks do not
directly affect informal finance, and also a proportion of borrowers alternate between formal and
informal finance.

The features highlighted above have inspired various studies with regard to monetary policy in
emerging economies. While some explored the operation of the traditional Keynesian interest
rate channel of the monetary policy transmission in a cross-section of developing country
contexts (Mukherjee and Bhattacharya, 2011), others, such as Mangani (2011), for example,
explored the presence of the various conventional channels of monetary policy in a single
developing country context, Malawi. The empirical results of Mukherjee and Bhattacharya,
(2011) suggest that private consumption and investment in the region are sensitive to
movements in real interest rates. Mangani (2011) found, for Malawi, that there was a lack of
unequivocal evidence in support of any of the conventional channels of the monetary policy
transmission mechanism, except for the exchange rate channel, which proved the most
important. Another recent study by Kovanen (2011) on Ghana also documents the interest rate
pass-through as being incomplete.

From their findings, the general theme seems to be that monetary policy transmission via
conventional routes such as bond markets and banking systems is weak in EMEs (Mishra et al.,
2011). This, they believe, flows from the fact that trade liberalization has run ahead of financial
liberalization in most emerging economies, thereby providing a much more responsive route for
monetary transmission via the exchange market.

3.2.3 The Exchange Rate Channel of Monetary Transmission

According to Frenkel and Taylor (2006), an appropriate level of the real exchange rate can be a
key element in the quest for growth, employment creation and overall development of the ‘real’
economy. This presupposes that knowing how to control the real exchange rate will have very
important implications for growth and employment. It is for this reason that several studies have

41
sought to unravel the exchange rate pass-through of monetary policy, also a key element of
macroeconomic policy. According to this school of thought, a contractionary monetary policy
stance stimulates appreciation of the domestic currency relative to foreign trading counterparts
thus making imports cheaper, switching demand away from domestically produced goods and
services and causing domestic price levels to drop.

As put forward by Adolfson (2001), in an open economy, inflation is greatly influenced by how
its determinants adjust to exchange rate movements. Mishkin (2007) also argues that when the
real price of imports fall along with domestic prices, monetary policy effects to the economy can
be captured as operating through the exchange rate effects on net exports.

From the literature (see Zorzi et al., 2007; Raveena and Natalucci, 2008), it seems the exchange
rate pass-through provides the most viable route for the transmission of monetary policy as
regards small, open or emerging economies. The importance of the exchange rate channel as the
most effective route for monetary policy transmission is predicated on the ineffectiveness of the
interest rate channel, which requires a well-developed financial sector to operate, a feature that is
usually non-existent in developing or emerging economies. In light of this, evidence on the
extent of the exchange rate pass-through is quite influential in the design of monetary policy
(Monacelli, 2005; Smets and Wouters, 2002; Adolfson, 2001; Devereux, 2001). This is because an
incomplete exchange rate pass-through for imports and domestic prices, according to Smets and
Wouters (2002), reduces the effectiveness of monetary policy via the exchange rate channel.

3.2.4 Determinants of Exchange Market Pressure

The literature on the determinants of exchange market pressure is scanty, as previous studies on
the exchange market have mostly employed changes in the nominal or real exchange rate to
proxy currency pressure. Given that EMP is an improved proxy for pressure on the exchange
market, its determinants are not fundamentally different from the determinants of nominal or
real exchange rates. For the purposes of this study, the following determinants are reviewed: the
monetary policy rate; output growth; private capital flows; current account balance; terms of
trade; and public debt.

Monetary policy rate: The central bank policy rate, a de facto indicator of a country’s policy
stance (see Tanner, 2001; Kamaly and Erbil, 2000; Gochoco- Bautista and Bautista, 2005), is
mostly employed as the proxy for monetary policy in studies on EMP. Increases in interest rates,

42
relative to trading partners, make investment in the higher interest rate economy more attractive
and directs the flow of capital into the economy. This inflow results in the excess supply of
foreign reserves thereby reducing exchange market pressure. This suggests that a contractionary
monetary policy stance (high interest rates) results in a reduction in EMP. Godfajn and Baig
(1998), Tanner (2001) and Gochoco-Bautista and Baustista (2005) are among the studies that
have sought to test this relationship empirically. They all find evidence to support the posited
inverse relationship between the monetary policy rate and EMP.

An alternate view, however, holds that increases in domestic interest rates may actually have a
perverse effect, thereby resulting in currency depreciation (Garcia and Malet, 2007). This view
argues that an increase in the monetary policy rate will trigger a rise in cost of capital, ultimately
reducing output growth. The fall in returns to investment and domestic output reduces demand
for the domestic currency and causes the currency to depreciate. This argument is further
buttressed by the notion that an increase in interest rates can also cause panic about expected
and actual bankruptcies among investors who may view the monetary policy stance as an
indication of the monetary authorities’ devaluation expectations. These arguments make for an
ambiguous impact of the monetary policy stance on EMP.

Output growth: According to conventional economic theory, GDP growth promotes currency
appreciation, all things being equal. As per the intuition, lower GDP growth generally implies
smaller demand for money (due to drop in demand), leading to expectations of currency
devaluation, and ultimately mounting pressure on the domestic currency. This suggests that a
country with strong economic growth will attract investment capital (for domestic assets) seeking
to earn higher returns. This capital inflow leads to higher demand for the domestic currency
relative to foreign currency and generates appreciation pressure (Balassa, 1964)1. Conversely,
nations with weak growth rates are likely to experience capital flight and weaker currencies, as a
result of demand for foreign denominated assets earning relatively higher returns. A number of
studies have included real GDP growth in studying EMP. Connolly and da Silveira (1977)
studied Brazil over a sample period of 1955 to 1975 and reported that GDP has a negative and
significant effect on the EMP. Hegerty (2010) also reports for Latin America, that output growth
eases EMP. Based on the available empirical literature, the conventional theory that growing
economies will experience appreciating currencies is supported.

1
Balassa (1964) argues that richer countries experience appreciating currencies.
43
Private capital flows: Flowing from the argument above linking capital flows to EMP, private
foreign capital flows are also included as an explanatory variable. The investment of private
capital inflows into infrastructural developments results in increased productivity and aggregate
demand. This translates into increased demand for the domestic currency and ultimately,
currency appreciation. On the other hand, where the capital inflows end up in the consumption
basket, it only serves to increase the relative price of domestic goods, increases the demand for
imported goods and spurs depreciation pressure (Combes et al., 2010). A number of other
studies have provided further distinctions to this argument by documenting that the impact of
private capital flows on the currency position and related adjustments depends on whether the
flows are pro-cyclical or counter-cyclical. According to Lueth and Ruiz-Arranz (2007) and Chami
et al., (2008), counter-cyclical flows of private capital only serve as buffers to the economy and,
therefore, have no influence on the exchange rate setup. They proceeded to indicate that pro-
cyclical flows have a tendency to overheat the economy and cause currencies to appreciate.

Despite the seemingly well-defined route through which private capital flows may affect the
currency management setup of any economy, the empirical evidence is still ambiguous. Lartey
(2007), for example, reports that aggregate capital flows are inconsequential in influencing the
exchange rate setup, using data on economies from Sub-Saharan Africa. In an earlier study on six
Central American Economies, Rajan and Subramanian (2005) also reported that aggregate capital
flows are insignificant in explaining the exchange market pressure of an economy (see also
Izquierdo and Montiel, 2006). Saborowski (2009), on the other hand, contradicts these findings
with results that suggest that private capital inflows to developing countries are significant in
explaining appreciation in the domestic currency. These arguments make for an ambiguous role
for private capital flows.

Current account balance: According to Rosenberg (1996), trade flows drive currency demand.
A positive or surplus trade balance on the current account implies an increase in the demand for
the domestic currency relative to the foreign currency. This increases foreign reserves and
reduces pressure in the foreign exchange market2. Most empirical studies on EMP suggest an
inverse relationship between the current account balance and EMP (see Hegerty, 2010; Khan,

2
Put differently, a rise in the current account to GDP ratio is generally associated with large external capital inflows
that are intermediated by the domestic financial system and could facilitate asset prices and credit booms and
thereby increase the probability of a currency appreciation via improved output and investment return (Berg &
Pattillo, 1999)

44
2010). Karfakis and Moschos (1999) find that the current account balance has a negative impact
on the foreign exchange market pressure. Hegerty (2010) also finds that the current account
balance exerts the most significant influence on EMP as predicted by theory (see also Van Poeck
et al., 2007). Khan (2010) strengthens this position by also reporting that the current account
balance is significant not only in explaining the current level of EMP but also that of two
subsequent future periods. Licchetta (2009) also provides evidence on the role of external
balance sheet variables in determinants of currency crises in emerging markets and advanced
economies. He additionally found that the likelihood of a crisis is increased with a larger current
account deficit.

Terms of trade: According to Bagchi et al., (2004) and Bergvall (2004), a general increase in
terms of trade (TOT), results in more imports being exchanged for one unit of exports. Ceteris
paribus, this leads to an appreciation of a local currency. Terms of trade also capture the
elasticities of foreign demand for domestic goods and domestic demand for foreign goods. A
significant deterioration in the terms of trade, either from declining export values or increasing
import values, is likely to precipitate a depreciation of the local currency. This can either directly
stimulate export demand or indirectly switch domestic demand to domestic products so as to
minimise pressure on the domestic currency. This suggests that the appreciation in the domestic
currency is driven by the increase in demand for the domestic currency due to the rising price of
the country’s exports. Conversely, deteriorating terms of trade imply a need for more foreign
currency to fill the gap between export revenue and import expense. Based on the views above,
terms of trade is expected to alleviate exchange market pressure.

Public debt: According to the Ricardian equivalence principle, increases in the levels of debt
have the same implications as government financing via tax increases. It only causes rational
economic agents to increase savings today in order to maintain equivalence in the future. This
neutralizes the effects of government deficit financing via debt. Opponents to this theorem,
however, indicate that accumulation of public debt could undermine investor confidence and,
therefore, lead to capital flight resulting in depreciation pressure on the domestic currency
(Mandilaras and Bird; 2008). Another argument is that the use to which the accumulated debt is
put is the key to its effect on the foreign exchange market pressure. Public debt is employed in
models of EMP to capture fiscal effects. Mandilaras and Bird (2008) in a study on Latin America
find evidence to support their argument that rising debt levels increase EMP. In general, it seems
that the impact of debt on EMP flows more from the reduction in investor confidence due to

45
perceived unsustainability of the public debt levels and an increase in vulnerability to shocks such
as reversals in capital inflows (Eichengreen and Arteta, 2000). The foregoing thus implies that
levels of public debt can operate in both the positive and negative dimensions with regard to
EMP.

3.3 Methodology and Data

This section details the sample, data sources and model for the study. It also discusses the
conceptual framework on which the econometric model is built.

3.3.1 Sample and Source of Data

The study uses annual data on 20 Sub-Saharan African countries3 over a 20-year period from
1991 to 2010 (see Appendix B for data sources and variable measurement). The countries used
in the sample are based on the Sub-Saharan emerging economies identification, carried out by
Kehl (2007) and Radelet (2010). These analyses employed macro political and macroeconomic
indicators in identifying emerging and upcoming emerging (threshold/frontier) economies.
Performance above the international average on a significant portion of the elements for the past
decade or more was used as the selection criteria. The following parameters, culled from the
Africa Foreign Investment Survey 2003 by the United Nations Industrial Development
Organization (UNIDO), were the metrics for the classification by Kehl (2007) -: political
stability; economic stability; local market access; transparency; legal framework; skilled labour
availability infrastructure and government agency support. Radelet (2010) used policy changes,
the rise of democracy and economic performance as the selection criteria. Merging the output of
the two studies produced a set of emerging and upcoming emerging/frontier economies.
Availability of data, however, reduced the sample to 20 countries.

The data for the study were captured from the International Financial Statistics and are annual data
series. The data series include central bank policy/discount rate (as proxy for monetary policy),
international reserves, the domestic currency to US Dollar exchange rate, the real GDP growth,
the current account balance (as a percentage of GDP), the terms of trade (captured as the price
of exports to imports), private capital flows and public debt.

3
Benin; Botswana; Burkina Faso; Cape Verde; Gabon; Gambia; Ghana.; Lesotho; Madagascar; Mali; Mauritius;
Mozambique; Rwanda; Senegal; Seychelles; South Africa; Swaziland; Tanzania; Uganda and Zambia.

46
3.3.2 Model Motivation
Exchange rate management regimes fall into two broad categories, fixed or freely floating with a
third being a hybrid of the two and generally referred to as a managed float. Under the floating
regime, disequilibria within the money market of an open economy is usually offset via an
adjustment in the exchange rate of that currency, while under a fixed regime the adjustment is via
foreign reserves. The managed float entails either an adjustment of the exchange rate, a change in
foreign reserves or a combination of the two, a phenomenon generally captured as exchange
market pressure (henceforth EMP). According to Girton and Roper (1977), EMP can be defined
as the sum of exchange rate depreciation and change in foreign reserves that is required to
restore equilibrium to the domestic foreign exchange market. Large, an increasing or positive
EMP is generally defined as depreciation pressure whilst small, decreasing or negative values
indicate pressure for the domestic currency to appreciate.

Given that EMP is a composite measure of adjustments to restore equilibrium to the money
market, macroeconomic characteristics that exert influences on the equilibrium of the exchange
rate market can be expected to also influence exchange market pressure. On the basis of
Jayaraman and Chee-Keong’s (2008) conclusion that EMP is fuelled by macroeconomic
fundamentals and on the basis of the EMP literature (see Kibritcioglu et al., 1998; Tanner, 2001;
Gochoco-Bautista and Bautista, 2005 Van Poeck et al., 2007; and others) EMP, in this study, is
modelled as being a function of the central bank policy rate, the rate of real GDP growth, the
current account balance, the level of private capital flows, the level of public debt and the terms
of trade. This is on the assumption that, ceteris paribus, changes in these macroeconomic variables
will influence the demand and supply of domestic currencies and thus affect the pressure that is
exerted on a country’s currency relative to major trading currencies. The general model for the
study is therefore stated as:

……………………………………………………………… (1)

Where EMPit is the measure of the exchange market pressure on a currency. It is captured as the
sum of the percentage change in the domestic price of the US Dollar (it is the currency most
countries trade in) and the changes in international reserves. In terms of interpretation, the
general tendency is to infer depreciation pressure when the EMP value is positive (increasing),
and to infer an appreciation pressure when the value is negative (decreasing). MPR it is the central

47
bank policy rate, a de factor measure of the monetary policy stance. RGDPit is the annual rate of
growth of GDP. CABAit is the current account balance as a percentage of GDP . PKFLit is
private capital inflows as a percentage of GDP. PDit is the stock of public debt as a percentage of
the GDP. TOTit is the terms of trade, proxied by the ratio of exports to imports. Ɛit is the error
term, with i and t representing country i in time t.

3.3.4 Estimation Methodology

Panel data regressions were used to test the relationship between exchange market pressure, the
central bank monetary policy rate, the current account balance, output growth, terms of trade,
private capital flows and debt. The use of panel data regressions allow for the control of omitted
variable bias that may result from the exclusion of unobservable variables. The increase in
number of observations, as a result of the panel data technique, also provides more information
and hence precision while allowing for the analysis of economic questions that cannot be
addressed using time series or cross-sectional data (Hsiao, 2003).

The dependent variable for the study is based on the EMP index of Girton and Roper (1977) for
capturing exchange market dynamics. The exchange market pressure index is modelled as a
function of the macroeconomic fundamentals. Given the possibility of persistence flowing into
the determination of EMP, dynamic panel data estimation is used, where the lag of the
dependent variable is also included as an explanatory variable. The system is estimated as:

EMPit   0  1 EMPit 1   2 MPRit   3GDPit   4 PDit   5 PKFLit   6TOTit   7 CABAit   it

(2)
where EMPit-1 is the lag of the dependent variable along with the policy rate, real GDP growth,
debt, private capital flows, the terms of trade and the current account balance.
The Arellano and Bond (1991) approach to dynamic panel estimation, using the difference
Generalised Method of Moments (GMM), is used. This is on account of the fact that the right-
hand variables cannot be said to be strictly exogenous, coupled with the possibility of
unobserved heterogeneity. Another reason is that the panel is unbalanced and the sample
selection is not random.

48
3.4 Results and Discussions
This section presents the discussion of study findings. The first section presents the correlation
matrix between the variables under study and the second section presents the regression output
for the various equations.

3.4.1 Correlation Analysis

In testing for the possible presence of multicollinearity in the regressors, a correlation matrix was
constructed. From Table 3.1, real GDP growth, private capital flows, and the monetary policy
rate are positively correlated with EMP. On the other hand, levels of public debt, the terms of
trade and the current account balance are negatively correlated with EMP. None of these, except
the terms of trade variable, is significant at 10%. The monetary policy rate is significantly
negatively related to terms of trade and the current account balance, but positively and
significantly related to debt levels. Private capital flows and output growth are negatively
correlated with the monetary policy rate, but the relationships are not statistically significant.
Private capital flow(s) is significantly and negatively correlated to terms of trade and the current
account balance, but positively significant to public debt levels, though its relationship with
output growth is insignificant.

49
 Table 3.1: Correlation matrix of study variables

EMP MPR PKFL TOT DEBT GDP CABA

EMP 1.0000

MPR 0.0634 1.0000

PKFL 0.0414 -0.0336 1.0000

TOT -0.0833* -0.1422** -0.1408** 1.0000

DEBT -0.0316 0.4668*** 0.1398** -0.2414*** 1.0000

GDP 0.0414 -0.0415 0.0490 -0.1206** -0.1251** 1.0000

CABA -0.0500 -0.1377** -0.4344*** 0.6508*** -0.3291*** -0.0900* 1.0000

***, **, *, = significance at 1%, 5% and 10% respectively. EMP is exchange market pressure; MPR is monetary policy rate; PKFL is private capital flow; TOT is terms of trade; DEBT is stock of public
debt; GDP is real GDP growth; CABA is current account balance.

50
Terms of trade is negatively and significantly correlated with both debt levels and real GDP
growth, but significantly positively related to the current account balance. Public debt is in turn
negatively and significantly correlated with output growth and the current account balance, while
output growth is significantly negatively correlated to the current account balance. Given the
high and significant correlation between the terms of trade and the current account balance,
these were alternated in the regression estimations so as to avoid multicollinearity.

3.4.2 Regression Results

The regression results are presented in Table 3.2. Due to the relatively high correlation between
the current account balance (CABA) and terms of trade (TOT), they are alternated in the
regressions (regression #1 and #2). As such the two do not feature simultaneously in each
regression. From the table, the lag of the dependent variable, EMP(-1), returns a negative and
significant coefficient. This indicates that past (higher) levels of EMP reduce current EMP and
suggests the absence of EMP persistence. It could be an indication that high periods of EMP are
usually expected to taper out and therefore end up doing so. The central bank policy rate, MPR,
has a significant and negative impact on EMP, which is consistent with Tanner (2001) and
Kamaly and Erbil (2002). It also confirms the position of the conventional argument, that there
is an appreciative effect when domestic interest rates are increased. This implies, that a
contractionary monetary policy stance that translates into higher interest rates has a tendency to
ease depreciation pressure on the domestic currency. Given that the countries under study are
yet to experience currency crises, the findings of Gochoco-Bautista and Bautista (2005) may
prove more relevant to the SSA situation. In light of this, the study findings can only be expected
to hold during tranquil periods but not in crisis periods.

The stock of public debt has a significant and negative effect on EMP. This indicates that rising
levels of public debt is characterised by lowering levels of exchange market pressure. This
contradicts the generally accepted notion which supports a depreciation effect for high levels of
debt due to perceived unsustainability (see Eichengreen and Arteta, 2000). The appreciating
effect seems to indicate that accumulated debt finds its way into the production function of the
sample countries to stimulate output growth. This then translates into high demand for the local
currency (as a result of a high demand for goods and investments from the domestic economy).

Output growth, captured as real GDP growth, shows a significant and positive relationship with
exchange market pressure. This indicates that increasing real GDP is associated with

51
depreciating domestic currencies, which is contrary to the general theoretical and empirical
literature which posits an appreciation effect (see Balassa, 1964; Samuelson, 1964; Hegerty,
2010). This is similar to the findings of Gunsel et al., (2010) from a panel of Latin American and
Asian Economies. The results suggest that currency depreciation in the face of increasing GDP
is usually underpinned by excessive local demand for foreign goods. This creates an excess
demand for foreign currency over and above the domestic currency and triggers depreciation.
Alternatively, the finding is also supported by the inflation-induced depreciation argument.
According to this intuition, increases in aggregate spending and the resulting increases in GDP,
when not matched by the productive capacity of the economy in the tradable sector, will result in
inflation and a subsequent depreciation in the currency (see also Combes et al., 2011).

Table 3.2: Regression output including South Africa

Variable Regression #1 Regression #2

EMP(-1) -0.0314146 (0.000)*** -0.0300204 (0.000)***

MPR -5.936033 (0.000)*** -6.035031 (0.005)**

PKFL -7.185381 (0.160) -7.807155 (0.272)

DEBT -1.967312 (0.000)*** -2.044135 (0.000)***

GDP 6.166616 (0.008)** 5.68232 (0.035)**

TOT - -254.6899 (0.007)**

CABA -4.7200282 (0.006)** -

Wald Chi2, Prob. 1484.17 (0.0000)*** 1365.86 (0.0000)***

Arellano-Bond Test Order 1 -1.3107 (0.1900) -1.31153 (0.1898)
for Autocorrelation Order 2 -0.34128 (0.7329) -0.39653 (0.6917)
Sargan’s Test 14.76282 (1.0000) 14.38127 (1.0000)
***, ** and * denotes significance at 1%, 5% and 10% respectively. Values in parenthesis are the p-values. The following
diagnostic tests are run for the model: the Sargan test for over identification, the Arrelano-Bond test for autocorrelation and
the Wald Chi2 for model fit. The results indicate that the over identifying restrictions are valid, there is no first order or
second order autocorrelation and the model fits the data well.

The coefficient of terms of trade (TOT), as expected, has a significant and negative effect on
EMP. This implies that as terms of trade improves (export earnings flow in in excess of import
expenses), the supply of foreign currency exceeds that of the local currency, creating appreciation

52
pressure on the domestic currency. This is consistent with the findings of Combes et al., (2011),
who document that the possibility of currency crises is heightened when terms of trade
deteriorates. Kaminsky et al., (1998) have also reported that an economy’s terms of trade is a
leading indicator of currency crises.

The second regression, which included the current account balance, (CABA), and therefore
excluded TOT for issues of multicollinearity, returned similar findings for the lag of EMP, the
policy rate, the stock of public debt and the GDP. The additional finding from the alternate
regression flows from the current account balance which shows the theoretically posited sign and
significance. It can be explained that increases in the current account balance contribute to a
reduction in exchange market pressure on the domestic currency. This finding supports the
finding of Eichengreen et al., (1995), that improving current account positions have appreciating
effects on the domestic currency. In general, it can be expected that as countries within SSA
work to improve their current account positions, their domestic currencies will appreciate
relative to foreign trading partners.

Private capital flows, (PKFL), as shown in Table 3.2, prove inconsequential with regard to EMP
in both estimations. A possible explanation for the insignificance of private capital flows could
be illegal capital flight or capital flight due to an unstable economy. With such capital flight,
inflows of private capital only come in to offset the loss of capital, hence the neutrality. Another
explanation is that capital flows that do not enhance productivity in the tradable sector have
negligible effects on the foreign exchange market. On another hand, if private capital flows come
in to stabilise economies from foreign exchange losses due to macroeconomic shocks then its
effect is negligible. Another reason, following Lartey’s (2007) argument, has to do with the
individual effects of the components of private capital flows. Whilst the other components of
private capital exert appreciation pressure, FDI which is the largest component of private capital
flows in Africa, exerts a depreciation pressure, hence a neutral effect for PKFL. It could also be
the case of “hot money”, the incidence where there is an inflow of funds into the economy but
on a temporary basis only, just to take advantage of favourable interest rates in the economy.
The insignificant finding is in consonance with Lartey (2007), Izquierdo and Montiel (2006) and
Rajan and Subramanian (2005), who document that aggregate private capital flows is not
influential in explaining the exchange market pressure on a currency. Following the opinion of
Lueth and Ruiz-Arranz (2007) and Chami et al., (2008), it can also be said that private capital
flows are insignificant because they are counter-cyclical and therefore only serve to neutralize

53
whatever currency pressures may be in existence within the economy. The finding, however,
contradicts Saborowski (2009), who suggests that for developing countries, private capital
inflows are significant in explaining appreciation in the domestic currency.

Sensitivity analysis and robustness checks

To test for the possibility of biases in the coefficient estimates, the models were subjected to
robustness checks. The first test was to exclude South Africa (being the oldest and largest
emerging economy in the sample) so as to rule out any form of domination by South Africa.
From Table 3.2 and 3.3, there is no indication of significant differences from the previous
regression that included South Africa in the sample. On the whole, the variables retained their
signs and significance, giving no reason to infer any biases in the coefficient estimates as a result
of including South Africa in the panel.

Further tests for sensitivity to specification also included the addition of lags of GDP to the
main model (results in Appendix B). Upon the addition of one lag, both the current level and the
lag of the GDP became insignificant but maintained their signs.

Private capital flows, however, became significant with the expected negative sign. This implies
that an increase in private capital flows causes an appreciation in the domestic currency. This
finding is in consonance with the findings of Hegerty (2010). In essence, it appears that when
countries are able to direct the flow of foreign capital into the economy, the local currency will
appreciate. Given that EMP captures changes in the exchange rate and international reserve
positions, inflow of foreign capital ultimately translates into accumulated reserves, thus causing
currency appreciation. It can thus be expected that countries that pursue policies to direct the
flow of foreign private capital flows into the domestic economy are likely to experience currency
appreciation.

Table 3.3: Regression output excluding South Africa

Variable Regression #1 Regression #2

EMP(-1) -.0315269 (0.000)*** -.0305206 (0.000)***

MPR -6.180829 (0.000)*** -7.232136 (0.001)**
PKFL -6.786846 (0.177) -3.537157 (0.487)

54
DEBT -1.966323 (0.000)*** -1.941454 (0.000)***
GDP 6.245621 (0.006)** 5.684366 (0.032)**
TOT - -209.1492 (0.002)***
CABA -4.62803 (0.005)** -
Wald Chi2, Prob. 1454.79 (0.0000)*** 1324.97 (0.0000)***
Arellano-Bond Test Order 1 -1.3096 (0.1903) -1.3074 (0.1911)
for Autocorrelation
Order 2 -.35955 (0.7192) -.52336 (0.6007)
Sargan’s Test 14.06068 (1.0000) 13.60909 (1.0000)
*, **, *** indicate significance at 10%, 5% and 1% respectively. The following diagnostic tests are run for the model: the Sargan test for
over identification, the Arrelano-Bond test for autocorrelation and the Wald Chi2 for model fit. The results indicate that the over
identifying restrictions are valid, there is no first order or second order autocorrelation and the model fits the data well.

This finding notwithstanding, the implication must be taken with caution since it only emerges
after the inclusion of the lag of GDP.

Tests for model suitability: All the regression models were subjected to the Arellano-Bond test
for autocorrelation and the Sargan’s test for over-identification respectively. The implausibly
perfect scores for the Sargan’s tests are not considered as indicative of mis-specifications given
that Abiad, Oomes and Ueda (2010) report similar p-values. On the whole, since both tests failed
to reject the null hypotheses, there is no reason to doubt the regression results generated.

3.5 Conclusion and Policy Implications

The study sought to test the hypothesis that contractionary monetary policy strengthens currency
positions (by reducing EMP), while controlling for other known macroeconomic determinants
of exchange market pressure. From the study, it emerged that the conventional argument of
defending a currency via increasing interest rates holds for the study sample. A rise in interest
rates is, therefore, more likely to trigger an appreciation of domestic currencies by redirecting
foreign inflows into the economy. Despite this finding, the position of Feridun (2009) that fragile
financial systems are prone to perverse reactions in the face of rising interest rates must not be
ignored completely (within such economies).

Contrary to the literature, debt is associated with appreciation, while economic growth is
associated with currency depreciation. Currency pressure, it seems, is also not persistent across
the region, with periods of high pressure being followed by periods of low exchange market

55
pressure and vice-versa. Terms of trade and the current account balance are highly significant in
explaining the currency pressure and have the posited appreciation effects. On the whole, it can
be said that while monetary policy, along with other macroeconomic fundamentals are relevant
in explaining currency positions, the effects of the fundamentals, for SSA, are not completely
consistent with trends in the existing literature.

In terms of policy implications, it would appear that within the SSA region, attempts at
addressing depreciation pressure on their domestic currencies can be achieved by increasing
interest rates. However, since these economies need lower interest rates to stimulate appropriate
levels of investment, this policy move must be implemented with caution. Improving the terms
of trade position and the current account balance will also provide a way of defending
speculative attacks on the domestic currency.

Another policy implication from the finding is with regard to the elasticity of domestic demand
for foreign goods. Policy moves aimed at stimulating demand in the direction of domestically
produced goods may hold the key to reversing the growth-currency depreciation cycle. This
could include export promotion policies in general through government investment in the
appropriate infrastructure. Also, support could be provided to target local industries to produce
goods that can compete favourably on both the domestic and the international markets. An
alternative, but more stringent approach may include the imposition of significant tariffs and
controls with regard to imported goods, especially consumables that are produced domestically.

56
 CHAPTER FOUR
MONETARY POLICY AND BANK RISK-TAKING BEHAVIOUR IN SUB-
SAHARAN AFRICA - SOME EMPIRICAL EVIDENCE

4.1 Introduction
Economies in search of high and sustainable levels of economic growth have, over the years,
been offered numerous prescriptions and possible solutions. One of such prescriptions, among
many, is the stimulation of capital accumulation via institutional and policy mechanisms
(endogenous growth theory: see Barro, 1990; Easterly and Rebelo, 1993 and others). According
to this theory, capital accumulation in the durable sector, driven by favourable policy
environments, has significantly positive influences on economic performance. In light of this
proposition, the role of monetary policy, for example, in stimulating capital accumulation and
ultimately economic growth has thus received significant research attention (Erceg and Levin,
2006; Amarasekara, 2008), with the general consensus being that a reduction in interest rates
spurs economic growth by stimulating capital accumulation via reduced costs of capital.

The empirical literature, following the financial crises of 2007 – 2009 has, however, questioned
the expansionary monetary policy-to-growth route. According to the literature, an expansionary
monetary policy, albeit capable of stimulating investment, may do so at the cost of excessive risk-
taking by financial intermediaries (see Borio and Zhu, 2008; Taylor, 2009; Agur and Demertzis,
2012). Borio and Zhu (2008) refer to this effect of monetary policy as the risk-taking channel of
monetary policy. As per this school of thought, banks, in their search for yield in low interest
rate settings, may take on higher yielding but riskier projects, ultimately pre-disposing them to
instability. Supporting this school of thought are Adrian and Shin (2009), who argue that as a
result of the wealth effect created by declining interest rates in an economy, previously riskier
firms that were unable to provide adequate collateral values to back required loans may
consequently be in a position to do so due to increases in their collateral (asset) values.
According to them, such a situation has the potential of putting the financial system on the brink
of instability in the event of negative shocks to these economic agents financed by the banking
system.

In the Sub-Saharan Africa (SSA), there is a quest to attain and sustain high rates of economic
growth. This quest for growth, according to Hernandez-Cata (2000), has created a need for
significant capital accumulation. In view of this need for significant capital accumulation, the

57
delicate balance between expansionary monetary policy, growth and financial stability
relationship raises very important policy and academic research questions. For instance, can SSA
pursue expansionary monetary policy to spur growth without the risk of financial instability?
And if SSA banks are prone to risk-taking when interest rates are low, is there another route to
growth that circumvents this path? Are there any bank-level or macro-level elements that may
worsen or mitigate this risk-taking behaviour? Risk-taking behaviour, for the purposes of this
study is defined as risk-taking that compromises the solvency of the bank.

While the literature is growing in volume and depth for most of the developed world in response
to some of these questions (see Jimenez et al., 2008; Ioannidou et al., 2009; Altunabas et al., 2010;
Maddaloni and Peydro, 2011), the same cannot be said for SSA. These questions are definitely
pertinent for SSA as well, and probably more so because of the dominant role that banks play in
the region’s financial system. So while the region has yet to experience a financial crisis on the
scale and magnitude experienced by the US, Europe, Asia and the Americas, it does not preclude
SSA from a study of the phenomenon, that periods of loose monetary policy in the form of low
interest rates can come with negative and undesirable effects like excessive bank risk-taking
behaviour and financial instability. In view of the above arguments, this study seeks to test
whether there is a risk-taking channel of monetary policy in Sub-Saharan Africa; how effective it
is; and whether there are bank-level and macro-level dynamics that mitigate or magnify the
impact of this channel. Indeed empirical answers to such questions hold relevant policy
implications for central bank monetary policy formulation as well as for macro-prudential
regulation.

In general, the study contributes to the literature in several ways: it seeks, first and foremost, to
document the possible existence of a bank-risk-taking channel of monetary policy in SSA. It also
seeks to empirically document possible determinants of risk-taking behaviour of banks in SSA.
In these regards, it also provides prudential and supervisory authorities with relevant insights that
will inform the level of vigilance that may be required within the financial system, at what times
and with respect to which institutions or to which economic agents.

The rest of the chapter is organized as follows: the next section reviews the literature on bank
risk-taking behaviour followed by the methodology in the next section. Sections four and five
present the study findings and conclusion respectively.

58
4.2 Literature Review

4.2.1 The Risk-Taking Channel of Monetary Policy

In the search for answers to explain the 2007 global financial crises that originated from the US,
a line of research has emerged that seems to suggest that there is a significant link between
monetary policy of low rates of interest for extended periods of time and high risk-taking by
financial intermediaries. This monetary policy effect was labelled the risk-taking channel of
monetary policy transmission by Borio and Zhu (2008).

According to the proponents of this ‘newly found’ transmission mechanism, its operation can be
delineated via three main routes. The first route is labelled the search for yield (see Rajan, 2005)
and operates in this manner: in periods of low interest rates, banks face a reduction in margins
between lending and deposit rates. This raises their incentives to switch to riskier projects with
expected higher yields. In so doing, their risk positions deteriorate. In the same framework,
banks may search for higher yields when the probability of defaulting on their existing long-term
liabilities is exacerbated by low yielding risky portfolios due to periods of low interest rates.

The second route under the risk-taking channel of monetary policy operates through the
expectations channel: expectations of default with regards to borrowers. According to Borio and
Zhu (2008) and Adrian and Shin (2010), low interest rates have the potential of inducing banks
to take on more risk because of favourable estimates of borrower default due to increasing asset
and collateral values. Put simply, in periods of low interest rates, economic agents that hitherto
were considered risky by virtue of inability to provide collateral now qualify for loans due to the
appreciation in collateral values and the reduction in price volatility. According to Borio et al.,
(2001) such reduction in risk perception also increases risk-tolerance and increases risky
behaviour by banks.

The third route, like the second, also works through expectations, this time expectations of a
continued decline in interest rates and of the level of economic activity. As per this line of
thought, when financial intermediaries anticipate rate reductions, they adjust their portfolio of
long-term assets towards riskier but higher yielding assets and especially so if they expect interest
rates to be kept at low levels for an extended or prolonged period of time (Agur and Demertzis,
2010; Valencia, 2011).

59
In the wake of the 2007 financial crises, the link between monetary policy and bank risk-taking
behaviour came up to the forefront of policy and academic discussions and so did the scantiness
of the literature in that area. The dearth in literature was attributable to the fact that past studies
on risk-taking by financial intermediaries had largely ignored the impact of monetary policy
(Dell’Ariccia et al., 2010) and the ample literature that had actually investigated the monetary
policy effects on credit markets had also largely ignored the risk-taking channel. Where monetary
policy had been incorporated, the focus had been on borrowers and quantity of credit (see
Huang, 2003; Hülsewig et al., 2006; Iacoviello and Minetti, 2008; Wu et al., 2011), but not on
lenders and on the quality of credit. So while the empirical studies were few, the arguments
suggest that persistently low interest rates fuelled booms in asset prices and securitized credit and
led to financial intermediaries taking on increased risk.

One of such studies, by Dell’Ariccia and Marquez (2006) finds the existence of a risk-taking
channel in periods of easy monetary policy and in the presence of information asymmetrical.
They report, that in an era of reduced cost of funding flowing from low interest rates, the
incentive to screen borrowers, which is costly and negatively affects profit, is greatly reduced,
leading banks to lend to riskier borrowers. This finding was also confirmed by Lown and
Morgan (2006) with their focus on the US.

After the financial crises, Maddaloni and Peydró (2010) also found evidence of risky behaviour
by banks during periods of low interest rates. They documented, for the Euro area, that credit
standards tighten when monetary policy takes a contractionary stance and loosen when monetary
policy is expansionary. Jiménez et al. (2008), using data on individual Spanish banks, found that
in periods of low interest rates, new loans have riskier profiles because banks lend to economic
agents with previous bad credit history. On the other hand, outstanding loans with variable rates
experience an improvement in credit risk because of the reduced rates and of the favourable
estimation of default probabilities.

Using Bolivian data, Ioannidou et al’s (2015) findings seem to indicate that when interest rates are
low, banks tend to reduce loan rates of risky borrowers, relative to their less risky counterparts.
This ultimately results in a higher risk profile for the bank. Further supporting evidence on the
existence of a risk-taking channel of monetary policy is provided by Altunabas et al., (2010).
According to their study findings, unusually low interest rates over an extended period of time
contributed to an increase in risk-taking for a set of listed banks across industrialized countries.

60
This seems to suggest that when interest rates are low, but for short periods only, risk-taking
behaviour among banks may not necessarily increase or become evident.

On the whole, the growing evidence seems to point to the existence of a risk-taking channel of
monetary policy. Notwithstanding the fact that the empirical evidence on the risk-taking channel
of monetary policy is growing in volume and in depth, the focus is mostly on developed markets
with little focus on emerging economies and least of all on Sub-Saharan Africa.

4.2.2 Determinants of Bank Risk-Taking Behaviour

Several theoretical and empirical positions have, over the years, appeared in the literature to
explain the risk-taking behaviour of banks. The generally accepted position is that bank risk-
taking behaviour is a function of both micro and macro level variables (see Salas and Saurina,
2003; Gambacorta, 2009). On the micro-level, some of the determinants of bank risk behaviour
identified include bank size, bank capital, bank profitability, lending growth and degree of
securitization activity (Crouhy et al., 2001; Maddaloni and Peydro, 2010). On the macro level, the
visible determinants are documented as including macroeconomic performance (GDP), stock
market returns, as well as the regulatory and competitive environments (Keeley, 1990; Boyd and
De Nicolo, 2005; Berger et al., 2009; Boyd et al., 2009). For the purposes of this study, the
empirical determinants to be discussed are: size, profitability, liquidity, capital and lending growth
on the bank (micro) level and economic performance on the macro-level, on account of the twin
factors of availability of sufficient supporting empirical evidence and availability of data for the
anticipated study sample.

Bank Size
The theoretical effect of bank size on bank risk-taking behaviour is yet to reach a consensus.
This is in spite of the fact that it is generally accepted as an important determinant of bank risk-
taking behaviour. On one theoretical side is the ‘too-big-to-fail’ perspective. According to
Demirguc-Kunt and Huizinga (2010), operating under the perception that they are too big to fail,
large banks have usually been found to take on excessive risks and ultimately end up failing. The
banks engage in excessive risk-taking on account of their perceived ‘safety nets’. The safety nets
may sometimes flow from government support, because they cannot be allowed to fail for fear
of destabilising the economy. On the other side is the diversification group. As per their
argument, larger banks, by virtue of the diversification effect, are inherently less risky. Put
simply, it is believed larger banks are less risky because they have greater potential to diversify
and therefore reduce risk-taking behaviour (see Hughes et al.,, 2001; Laeven and Levine, 2006

61
and Altunabas et al., 2010). Critics of the ‘diversification-reduced risk-taking’ school, however,
believe that large banks actually do take advantage of their diversification benefits to engage in
more risky behaviour (Keeley, 1990; Allen and Gale, 2000). Others, such as Scharfstein and Stein
(2000) and Goetz (2011), also believe that the increase in risk-taking behaviour of diversified
banks could flow from intensified agency problems linked to size.

On the empirical level, the research findings seem to mostly indicate that big banks do take on
excessive risks, either operating along the ‘too-big-to-fail’ paradigm or taking advantage of their
diversification benefits to engage in risky behaviour. A study by Demsetz and Strahan (1997)
concludes that banks use their diversification benefits to increase risk-taking behaviour.
Gonzales (2005) also supports the finding with evidence that large banks actually operate under
the ‘too-big-to-fail’ mind-set and end up taking on more risk. Deelchand and Padgett (2009),
using data on Japanese co-operative banks also document the same finding. Chen et al., (2006)
and Mehran and Rosenberg (2008) go a step further to disaggregate bank risk into systematic and
asset specific risk. Their findings also show that while size does affect risk-taking, the effect is
negative for asset-specific risk but positive for systematic risk. This finding seems to suggest that
while risk-taking behaviour may be a function of diversification, the perceived risk in their
profiles is more of systematic in nature than of an asset-specific nature (see an earlier study by
Diamond, 1984).

Bank Liquidity
Drehmann and Nikolaou (2010) define bank liquidity as the relative ability of a bank to honour
monetary obligations with immediacy and without undue stress. An improvement in a bank’s
liquidity position, by inference, is therefore an indication of sufficient funds to honour matured
liabilities and the ability to raise cash at short notice if needed (Borio, 2000). In terms of the
theoretical link between bank balance sheet liquidity and risk-taking behaviour, one view holds
that in periods of low interest rates excess liquidity proves costly for the bank (in view of liability
commitments with higher yields, compensation clauses, etc.). This may cause the bank to go in
search of higher yielding but riskier investments (Rajan, 2005) and cause deterioration in its risk
profile. Repullo (2004) also develops a predictive model that indicates that when banks have
access to liquidity, even off-balance sheet liquidity from the Central bank (lender of last resort),
they are very likely to assume risk profiles beyond their internal liquidity means.

Ioannidou et al., (2009), for instance, find that liquid banks take on more risk by virtue of their
search for yield. Berger and Bouwman (2010) also confirm this finding with research that

62
concludes that a high occurrence of financial crises is usually preceded by high bank liquidity
creation. This gives further support to the study finding by Acharya and Naqvi (2012) that when
banks experience a liquidity influx, they expand lending with a disregard for downside risk,
ultimately under-pricing the risks of projects and in the process making bad quality loans. Myers
and Rajan (1998), with a contrary opinion, earlier argued that in times of low interest rates, banks
are rather capable of refinancing existing assets at lower rates thereby decreasing default
probabilities and reducing overall bank risk. From the foregoing, one cannot thus immediately
conclude on the effects of liquidity on risk-taking behaviour. With regard to the empirical
literature, though, the evidence points more to risk-taking in the presence of excess liquidity.

Bank Capital
The effect of bank capital on bank risk behaviour is believed to be dependent on whether or not
the bank is mostly equity-funded or debt-funded. It seems that banks that are highly leveraged
have a tendency to take on excessive risks because potential losses from such risky behaviour will
not be internalized. On the other hand, when there is significant equity at stake, bank risk taking
behaviour is constrained. This is because capital serves as a commitment device to limit risk-
taking, therefore less leverage leads to less risk-taking. On the basis of this argument, the
conventional standpoint seems to lean in favour of the idea that well-capitalized banks, with
higher equity at stake, take on less risk compared to their undercapitalized counterparts. Furlong
and Keeley (1990) provide some empirical findings to support the conventional argument that a
higher equity capital at stake reduces incentives for excessive risk taking by banks.

Kwan and Eisenbeis (1997), however, argue that banks hold on to more capital so they can take
on more risk. They also provide empirical evidence that seems to indicate that capital
requirements may have the perverse effect of inducing risk-taking behaviour. This seems to
suggest that well-capitalized banks are well-capitalized just by virtue of their risky asset profile
(see also Hellman, Murdock and Stiglitz, 2000). Shrieves and Dahl (1992) also find evidence that
even for banks that are not constrained by regulations, increases in capital is positively related to
risk-taking. Koehn and Santomero (1980) and Kim and Santomero (1988) suggest that the
increase in risk-taking behaviour by banks flows from a reconfiguration of the composition of
their risky asset portfolios, because they face a forced reduction in leverage. This implies that due
to the increased regulatory capital requirements, banks may choose riskier asset profiles to make
up for the loss of utility originally derived from using debt capital. Iwatsubo (2007), apparently
on middle ground, argues that the relationship between bank capital and risk is actually non-
linear. Agoraki et al., (2011), on the other hand, hold the view that while capital requirements may

63
reduce bank risk behaviour in general, the effect is insignificant or even reversible with regard to
banks that have high market power. On the whole the empirical evidence provides no conclusive
evidence on the risk taking effects of bank capital, thus leaving it open as an empirical question
with an ambiguous effect.

Bank Profitability
The school of thought on the risk-taking effects of bank profitability argues that poor
performing banks have a tendency to pursue risky investments, the incentive being to re-
establish profitability – the so-called ‘gambling-for-resurrection’ effect. This argument is
intrinsically supported by Rajan’s (2005) search for yield proposition. In essence, when a bank is
faced with declining profits due to internal or external shocks, its search for higher yields may
ultimately result in risky behaviour. Delis and Kouretas (2011) contradict this argument with a
hypothesis that banks may ultimately show higher profitability on account of high-risk assets in
their portfolio and subsequently use the previous period’s profit to generate new and risky assets.
On the other hand, if their risky assets run into problems (bad loans), the resulting drop in
profitability will trigger a drop in the bank’s risky asset portfolio. In essence, they suggest that
higher not lower profitability drives increased risky behaviour and vice-versa. In this regard, they
suggest that the effect of profits on risk taking is with a lag. Delis and Kouretas (2011), however,
do not find empirical evidence to support the increased profitability to increased risk taking
hypothesis. Other studies have, however, provided some empirical evidence for the declining
profits to risky behaviour argument: Allen and Gale (2004), Claessens and Laeven (2004) and
Hellman et al., (2002) among others. According to Allen and Gale (2004), reduction in monopoly
rents and ultimately profits, as a result of competition, may force banks to engage in risky
behaviour to garner more profits. Claessens and Laeven (2004) narrow it down to competition
that flows from unconstrained banking activities leading to erosion of profits and thus inducing
risky behaviour. Hellman et al., (2002) also present evidence similar to Claessens and Laeven
(2004). By implication, prior losses may be the motivator for risky bank behaviour in subsequent
periods.

Excess Lending Growth

Lending, the principal activity in financial intermediation, is mainly the management of risks in
the pursuit of profits. The most significant of these risks is credit risk. Excessive lending growth,
by inference therefore, may be an a priori indicator of a bank’s risk taking behaviour. On one
hand, lending growth may be a reflection of improved borrower capacity. On the other hand, it
may also reflect banks’ engagement in morally hazardous behaviour or the loosening of credit

64
standards, with these two contributing the most to bank risk. The available empirical evidence
also seems to suggest that rapid expansion of lending by banks often leads to poor quality loans
because the growth in lending may exceed the lender’s capacity to properly evaluate and monitor
borrowers.

Reinhart and Rogoff (2009), for instance, indicate that most systemic banking crises have been
preceded by periods of excessive lending growth. Dell’Ariccia and Marquez (2006) and Tornell
and Westermann (2002) earlier argued that aggressive lending strategies have been usually
associated with risk concentration. Altunabas et al., (2010) later documented a non-linear
relationship between lending growth and bank risk. The findings of Altunabas et al., (2010)
suggest that when a bank’s lending, at the first instance, is too small to benefit from economies
of scale, the lending bank is ultimately riskier. This risk tapers off when economies of scale
benefits exceed the costs. When banks continue with aggressive loan growth, they once again
end up with riskier positions. Agur and Demertzis (2012) suggest that aggressive lending growth
can translate into risky behaviour for banks if bank loan officers are compensated on the growth
rate of their risky portfolios, but not necessarily on the profitability of such assets. This
compensation framework results in a disregard for downside risks in the pursuit of business.
Gambacorta (2009) also argues that aggressive lending growth is linked to riskier banks, using an
international dataset of listed banks across the European Union and the US.

Economic Growth
The role of macroeconomic performance in bank risk-taking behaviour, according to Matsuyama
(2007), works via economic expectations of lenders about borrowers’ net worth. The study
argues that increases in borrowers’ collateral values improve the overall perception of
creditworthiness of the borrowers by banks. This gives banks greater incentives to increase
lending and ease financial constraints, thereby taking on more risks. According to Kashyap et al.,
(1993), “better economic conditions increase the number of projects becoming profitable in
terms of expected net present value”, thereby influencing the perception of downside risks by
banks, and leading to more risk taking behaviour. On the other hand, better economic
conditions can contribute to a reduction in the proportion of non-performing loans on the
banks’ balance sheets, thereby contributing to a reduction in a bank’s risk profile. From the
foregoing, there seems to be conflicting positions on the impact of economic performance on
bank risk taking behaviour.

65
Camara et al., (2012) find empirical evidence to support the two conflicting positions. Using data
on European banks for the period 1992 – 2006, they found that while better economic
conditions, measured by the real GDP growth, reduced the credit risk exposure, the expectations
attributable to the booming economy also fostered risky behaviour by banks, thus skewing the
composition of their loan portfolios towards riskier projects. Dell’Ariccia et al., (2009), however,
provide evidence to support the argument that when the economic outlook is good, banks tend
to provide more credit, reduce their credit standards and ultimately increase their risk. Delis and
Kouretas (2011) provide further support for Dell’Ariccia et al., (2009) by documenting a similar
nexus between GDP growth and risk, using data on the European banking sector.

4.3 Methodology and Data

This section details the sample, data sources and model for the study. It also discusses the
construction of the study variables.

Sample and Source of Data. The study employed a total sample of 91 commercial banks from
12 emerging economies in Sub-Saharan Africa. These countries were chosen mainly because of
data availability and also because they have deregulated their financial sectors to foster financial
development and growth, hence providing an appropriate setting to test the behaviour of market
agents in response to monetary policy actions. Annual financial data spanning 1999 to 2012,
derived from BankScope were used. The panel data was unbalanced.

Model Motivation and Estimation. Based on the literature on determinants of bank risk
behaviour and the risk-taking channel of monetary policy previously discussed, bank risk
behaviour was modelled as being dependent on monetary policy, bank-specific variables and
other macroeconomic controls. The empirical specification (equation 4.1) follows Gambacorta
and Mistrulli (2004) and is designed to test whether the characteristics of the banks cause them
to react differently to policy changes, given relevant controls. Delis and Kouretas (2011) further
this model specification with the inclusion of interaction terms to test whether the total effects
on interest rates changes with bank characteristics. This study employs a similar approach.

Given that risk is a potentially persistent phenomenon on account of temporal smoothing,

competition and regulations, a dynamic panel is employed where a lagged dependent variable is
also included as a determinant. To avoid the possibility of an endogeneity bias as regards the
other study variables, the determinants are all lagged following Kashyap and Stein (1995) and
Altunabas et al., (2010). The general testable model is thus represented below, following

66
Altunabas et al., (2010) and Borio and Zhu (2008) on the linear assumptions underlying bank risk
and its determinants.

(4.1)

Where is the measure of bank risk-taking behaviour proxied by the Z-score,

is the monetary policy stance, is growth in real GDP, is for bank size, is for bank
liquidity, is excess growth in bank lending, is capital, is bank profit and
represents bank i in time t.

The study uses micro data because it allowed for the minimization of systemic effects while
allowing for the control of bank specific effects that might impact risk-taking behaviour. The
macro level is also controlled for on account of the fact that intermediation by banks is done
mainly towards residents, as such it is expected that country-specific variables may also play an
influential role as far as bank behaviour is concerned.

4.3.1 Construction of Study Variables

Bank Risk.
Bank risk-taking, is measured using the z-score of each bank. The z-score is
computed as the return on assets plus the capital-asset ratio scaled by the standard deviation of
asset returns. The z-score is a measure of bank stability and indicates the distance from
insolvency, where insolvency is defined as a state where losses exceed equity. A number of
studies on bank risk have employed the z-score (see Mercieca et al., 2007; Demirguc-Kunt and
Huizinga, 2010 and Laeven and Levine, 2009) In fact, according to Laeven and Majnoni (2003)
and Bikker and Metzemakers (2005) risk proxies like loan loss provisions and non-performing
loans are traditionally backward looking and pro-cyclical, hence the argument for the use of the
z-score. Boyd et al., (2009) define the z-score as indicating “the number of standard deviations
that a bank’s return on assets has to drop below its expected value before its equity is depleted
and the bank becomes insolvent”. According to Boyd et al., (2009), it is a measure of risk that is
“monotonically associated with the probability of bank failure” and also widely used in the
finance literature. De Nocolo et al., (2004) also indicate that this risk measure increases with
higher profitability and levels of capitalization but decreases when earnings become unstable,

67
inferred from higher standard deviation of return on assets (ROA). A higher z-score, intuitively,
therefore, indicates that the bank is more stable and engaging in less risky behaviour.

Monetary Policy.
The general banking literature is still far from settling on the best indicator of a country’s
monetary policy stance. A number of studies have employed the growth rates of monetary
aggregates or growth in credit. These approaches have, however, been criticised for being
ineffective in the face of innovation and deregulation (Bernanke and Mihov, 1998). The
monetary policy stance, , for this study is proxied for by the central bank policy rate/ prime
rate following Delis and Kouretas (2010). As per their study, the choice of interest rate measure
does not lend much difference to the outcome. This central bank rate is thus captured as the de
facto indicator of the monetary policy stance in the economy because it is the primary tool used
by central banks for implementing monetary policy. Based on the literature earlier discussed, a
low monetary policy rate is expected to increase bank risk-taking behaviour.

Economic Performance.
The macroeconomic environment is proxied by real GDP growth, . Aside from measuring
the level of economic activity, the GDP is also quite informative with regard to other
macroeconomic characteristics. On the country-level, it also addresses and controls for the
different levels of economic development in the countries under study. In view of the literature
available and reviewed (Camara et al., 2012; Delis and Kouretas, 2011 among others), a high level
of economic performance in a country can cause banks to favourably adjust their risk
perceptions and ultimately make riskier and more default-prone loans. On the other hand, an
improvement in the macroeconomic situation could be an indication of more profitable projects
within the economy and hence an improvement in the default risk of existing loans, which may
ultimately show up as a reduction in bank risky behaviour. Expectations on the sign are therefore
open.

Bank Size. Bank size, , is taken as the log of the bank’s total assets. It measures size
effect characteristics such as the bank’s market power, returns to scale, and diversification
benefits. On one side, size is expected to be related to bank risk negatively, owing to potential
diversification effects, or promote risk-taking behaviour on account of the too-big-to-fail
argument. Given these differing positions, the expectation with regard to size effect on bank risk
is also left open.

68
Bank Liquidity. Bank liquidity, , measures the funding liquidity position. It is calculated as
the proportion of cash and all short-term claims from other banks in relation to banks’ total
assets. It provides a relative measure of banks’ capacity to pay obligations with immediacy as well
as their ability to easily raise funds by liquidating short-term securities. Very low values may be
reflective of a bank with a liquidity problem, while unusually high values, except in instances of
high returns on riskless assets, may be indicative of a bank with no good lending prospects.
Either way, the liquidity may trigger risk-taking behaviour in two ways: as banks attempt to
offload their excess liquidity, there is a propensity to relax lending rules, thus accumulating risky
portfolios. On the other hand, a very low liquidity position may result in financial distress and
possible bankruptcy. The a priori expectation is that high liquidity which might trigger a search
for yield or low liquidity which may reflect liquidity crisis will reflect in a bank’s risk profile.

Bank Profits. Profitability, , measuring financial performance, is critical to a firm’s risk

analysis and indispensable to good decision-making. Profitability for this study is calculated as
profits before taxes, scaled by total assets. It invariably also provides a view into the cost
efficiencies (funding costs, operational costs) of the bank. Lower profitability can provide
incentives for banks to search for yields and thus engage in risky behaviour or it may trigger a
downsizing of risky portfolios so as to stall declining profits. On the basis of that, the effect of
bank profitability on risk-taking behaviour is left open.

Bank Capital. The level of bank capital, , is measured as the ratio of equity capital to total
assets. A relatively high value reflects the level of equity available to absorb losses, or on the
other hand, the level of equity provided as security to other players in the intermediation market,
such as creditors. The effect of capital on bank risk could go either way, either limiting risky
behaviour on account of the capital at stake, or taking on risk because of the capital buffer
available.

Lending Growth. Lending growth, , which will also reflect the level of credit
expansion within individual banks is captured as the year on year growth in credit for an
individual bank. This is captured as the difference between annual industry credit growth and
annual credit growth of individual banks, following Altunabas et al., (2010). Given that excessive
lending booms usually precede systemic crises (Reinhart and Rogoff, 2008; Borio and
Drehmann, 2009), controlling for this allows for the testing of an actual risk-taking channel of
monetary policy. Ceteris paribus, excess lending growth is expected to be linked to deteriorating
risk positions.

69
4.4 Findings and Discussions
This section presents the findings of the study along with the discussions. On account of the
larger panel and relatively small time series, the Arellano-Bond General Methods of Moments
(GMM) approach was employed in the dynamic panel estimation. The dynamic GMM panel
estimator was chosen because it simultaneously accounts for unobserved country specific effects,
allows for the exploitation of the time series variation in the data whilst controlling for the
presence of endogeneity in the explanatory variables. Table 4.1 below provides the correlation
matrix of the variables under study.

4.4.1 Correlation Analysis

From the correlation matrix, only bank profits show a positive relationship with bank risk
behaviour. Based on this preliminary relationship, it appears that high bank profits translate into
more stable banks. Both of the other variables of interest: monetary policy rate (MPR) and
liquidity (LIQ) show negative and significant correlations with bank risk-taking behaviour.
This seems to suggest that contractionary monetary policy (increasing policy rate) will likely
reflect in unstable banks in the financial sector. Increasing levels of bank liquidity also appear to
increase risky behaviour and reduce stability. RGDP growth, excess lending growth and bank
size, however, exhibit insignificant negative correlations with bank risk, probably indicative of
the fact that the economic setting, the bank’s size and lending behaviour do not affect bank
stability. On the other hand, there seem to be positive correlations between the monetary policy
rate, profits and bank liquidity. This may be evidence that as the policy rate increases, more
deposits are attracted into the banking sector while at the same time banks generate good profits
on loans made during these periods. The negative and significant correlation between the policy
rate and bank size may, however, temper the earlier view since it appears periods of
contractionary monetary policy can also lead to “shrinking” banks. Liquidity is positively related
to bank size and the real GDP growth rate while profits seem positively related to bank size and
real GDP growth. Excess lending growth was the only variable that was not correlated in any
significant way with any of the variables under study. None of the variables exhibited
correlations that were high enough to raise concerns of multicollinearity.

70
Table 4.1: Correlation matrix of study variables

ZSCORE MPR LIQ BSIZE ELNDGR PRFT CAP RGDP

ZSCORE 1.0000

MPR -0.0688
(0.0301)**
LIQ -0.1401 0.1302 1.0000
(0.0000)*** (0.0000)***
BSIZE -0.0180 -0.1046 0.1565 1.000
(0.5732) (0.0005)*** (0.0000)***
ELNDGR -0.0018 -0.0005 -0.0171 0.0053 1.000
(0.9545) (0.9870) (0.5713) (0.8602)
PRFT 0.0852 0.1532 0.0250 0.0625 0.0012 1.0000
(0.0067)** (0.0000)*** (0.4064) (0.0387)** (0.9694)
CAP 0.0852 -0.0332 -0.0313 -0.0632 -0.0008 0.2487 1.000
(0.0071)** (0.2696) (0.2989) (0.0362)** (0.9789) (0.0000)***
RGDP -0.0408 -0.0001 0.0763 0.0921 0.0002 0.0125 0.0161 1.0000
(0.2068) (0.9964) (0.0130)** (0.0027)*** (0.9941) (0.6833) (0.5979)
p-values in parenthesis; ***, **, * = significant at 1%, 5% and 10% respectively. ZSCORE is measure of bank risk; MPR is central bank policy rate and proxy for monetary policy; LIQ is bank liquidity;
BSIZE is bank size (log of total assets); ELNDGR is excess lending growth in the year; PRFT is bank profit; CAP is bank equity capital and RGDP is real GDP growth

71
4.4.2 Regression Results
Table 4.2 shows the dynamic panel regression output for equation 4.1. From the output (Table
4.2), the coefficient of the lag of the dependent variable is positive and significant, indicating
persistence in the risk behaviour of banks. The results seem to suggest that bank risk behaviour
in SSA is highly dependent on history. In other words, risky banks are likely to continue being
risky while stable banks will continue in their stability. The positive sign and significance of the
lagged dependent variable can also be viewed as the speed of convergence to the optimal risk
level with values closer to zero, implying a higher speed of adjustment. The results thus seem to
suggest that previous periods of risky behaviour are usually followed by less risk-taking in the
subsequent period.

The finding of this study appears to go against the findings of Agur and Demertzis (2012),
Taylor (2009) and Borio and Zhu (2008) and the bank risk-taking argument, which says that it is
low levels of interest rates that cause banks to go in search of yields and ultimately take on risky
positions. The coefficient of the monetary policy rate, MPR, is negative and significant. By
inference, increasing nominal policy rates reduces banks’ Z-SCORE. It thus appears that a high
central bank rate rather triggers risky behaviour among banks. On the flipside, however, it is no
secret that most of the countries in the sample are characterized by high inflationary
environments. The high inflation in these economies may still cause high policy rates to be
associated with low real interest rates and hence trigger risky behaviour by the banks. In this, the
risk-taking channel of monetary policy cannot be unequivocally rejected. Another possible
explanation could be that because deposit rates are generally low in SSA, as rates rise, deposits
dry up because savers go in search of better yielding investments. This shortage of cheap
deposits may force the banks into costlier funding sources and ultimately cause a search for high
yielding but riskier investments to balance out and maintain margins. It could also be that as
interest rates rise and filter through bank lending rates, only risky borrowers take up the high
interest rates hence the deterioration of bank risk profiles. These results, however, should be
interpreted with caution, as there may be other factors influencing the risk behaviour of banks,
like inflation, but have not been explicitly modelled in equation 4.1.

The coefficient of bank size (BSIZE) is negative and significant in relation to the Z-SCORE. It
indicates that the relationship between bank size (BSIZE) and risk-taking behaviour leans
towards the ‘too-big-to-fail’ argument, providing support for the studies of Dermiguc-Kunt and
Huizinga (2010), Deelchand and Padgett (2006) and Gonzales (2005). These studies found that

72
large banks engage in risky behaviour given their perception of available ‘safety nets’. The
findings of this study could also be indicative of the possibility that larger banks may be in
competition to capture even larger shares on the loan market (Stiglitz and Weiss, 1981),
following the ‘competition-fragility nexus’ and ultimately take on more risks, lured by the
prospects of higher returns.

The coefficient of excess bank liquidity, (LIQ), is also negative and significant. It appears that
excess liquidity seems to encourage risky positions in the banking industry. This is to be expected
given that unutilized excess liquidity, unless invested in high yielding government instruments,
exposes the bank to lower profitability. This finding supports the search-for-yield hypothesis by
Rajan (2005). The finding also supports the conclusions of Ioannidou et al., (2008) that liquid
banks take on more risks by virtue of their search for higher yields. The findings further support
the conclusions of Repullo’s (2004) predictive model that indicates that when banks have access
to liquidity, even off-balance sheet liquidity from the Central bank (lender of last resort), they are
very likely to assume risk profiles beyond their internal liquidity means. In essence, easy access to
liquidity, as posited by Acharya and Naqvi (2010), expands lending with a disregard for downside
risk, ultimately leading to an under-pricing of the risks of projects and in the process making
worse quality loans. This may suggest a desire for higher returns and larger market shares, thus
reducing the incentive to properly screen borrowers, and this culminating in higher risk profiles.

Table 4.2: Regression results for main model: Bank risk as a function of micro and
macro characteristics

Variable Regression #1

Zscore(-1) .213875 (0.000)***

MPR(-1) -0.5005776 (0.000)***
BSIZE(-1) -9.795589 (0.0000)***
LIQ(-1) -14.72113 (0.000)***
ELNDGR(-1) .0144758 (0.000)***
PRFT(-1) 23.3672 (0.000)***
CAP(-1) -41.55149 (0.000)***
RGDP(-1) .9071676 (0.000)***
_CONS 74.93821 (0.000)***

73
Wald Chi2 (Prob.) 61466.31 (0.0000)
Arellano-Bond Test Order 1 -3.0582 (0.0022)***
for Auto-correlation Order 2 -.82755 (0.4079)

Sargan’s Test 71.06733 (0.6689)

***, **, * = Significant at 1%, 5% and 10% respectively. Z-Score measures bank risk-taking; MPR is the central bank prime rate,
proxy for monetary policy; BSIZE is the log of total assets for size; LIQ is for bank liquidity; ELNDGR is excess lending growth;
PRFT is bank profitability; CAP is bank equity ratio; RGDP is real GDP growth. The following diagnostic tests are run for the
model: the Sargan test for overidentification, the Arrelano-Bond test for autocorrelation and the Wald Chi 2 for model fit. The results
indicate that the over identifying restrictions are valid, there is no first order or second order autocorrelation and the model fits the data
well.

Excess lending growth (ELNDGR), contrary to expectations, is positive and significant. This
suggests an improvement in the risk positions of the banks as a result of excess lending. It could
be an indication that excess bank lending still finds its way into safe and viable ventures, thus
contributing to healthier risk profiles of banks in the SSA region. The explanation could also be
that credit extension by banks in SSA is still limited so what emerges as excess lending above
industry standards is still significantly below the threshold that can trigger a deterioration in bank
solvency positions. In essence, what was captured as excess lending above the annual industry
average cannot be considered as excess lending, possibly an indication of credit constraints in
SSA. Another inference could be that since the SSA financial markets are less developed,
competition in the demand market for loans far outstrips the availability in the supply market.
This ultimately ensures that only creditworthy borrowers constitute the banks’ risky asset
portfolio.

Profit-making banks, from the results of the study, are ultimately less risky. This seems to go
against the standard high risk - high return hypothesis. A possible explanation for the positive
sign could be ‘cream skimming’ by banks in SSA. Flowing from the underdevelopment in the
financial markets in SSA, the market for bank assets is possibly large and populated enough to
make for profitable but relatively less risky banks. Another intuition could be that profitable
banks are profitable by virtue of their effectiveness in risk management, since banking is the
business of making profits by taking calculated risk. On another level, this finding seems to
suggest that losses in the prior year are a trigger for risky behaviour in the subsequent year. The
findings of the study reveal that less profitable banks are more likely to “gamble for

74
resurrection” after a period of losses and thus appear more risky. On another hand, higher profit
places the bank in a position to be able to absorb any external shocks, thus emerging as less risky
in its behaviour. The findings indirectly support the findings of Allen and Gale (2004) and
Claessens and Laeven (2004). According to these studies, low profits in a previous period trigger
risky behaviour in subsequent periods. This may suggest that high profits in previous periods are
likely to reduce risky behaviour in subsequent periods. A possible explanation could be that
banks experiencing good profits have no need to gamble for a resurrection by engaging in risky
behaviour.

The coefficient of bank capitalization is significant and negative. It is thus expected that well-
capitalized banks are more likely to engage in risky behaviour. This seems to support the
hypothesis that the well-capitalized are so capitalized because of their risk profiles. The finding,
however, goes contrary to the argument that well-capitalized banks are more cautious and
therefore less likely to take on risky positions since they have too much to lose from such risky
positions. In essence, the capitalized banks engage in more risky behaviours because there is a
buffer to absorb potential losses that may arise from the risky positions taken. The finding could
also be indicative of an attempt to offset the loss of utility derived from originally employing
leverage in its capital structure by investing in higher yielding but ultimately riskier asset
portfolios. This finding supports the findings of Kwan and Eisenbeis (1997), Kim and
Santomero (1988) and Hellman et al., (2000) that capital does actually trigger more risky
behaviour among banks.

Real GDP growth, the proxy for the macroeconomic environment, exhibits a positive
relationship with bank risk behaviour. It means that in periods of good macroeconomic
performance, bank risk behaviour declines. It thus appears from the results, that in a booming
economy, the banking sector exhibits a less risky profile by virtue of its portfolio reflecting more
profitable projects not prone to default. The findings, however, contradict Dell’Ariccia et al.,
(2010) and Delis and Kouretas’ (2009) findings that indicate an increase in bank risk behaviour
due to favourable perception of borrower risk profiles during economic booms.

4.4.3 Catalytic or Moderating Factors

Using interacting terms, the study also sought to find out the impact of bank-specific
characteristics on risk behaviour vis-à-vis the monetary policy environment. In this regard, three

75
characteristics: bank liquidity, bank capital and bank size were interacted with the prime rate. To
avoid significantly compromising the degrees of freedom for the study, the interactive terms
were included, one after the other and thus given as follows:

(4.2)

(4.3)

(4.4)

Where MPR*LIQ is the interaction between monetary policy and bank-level liquidity, MPR*CAP
is the interaction between monetary policy and bank capital and MPR*SIZE is the interaction
between monetary policy and size of bank.
The regression results in Table 4.3 show the impact of interacting bank liquidity with the policy
rate (Equation 4.2).

Table 4.3: GMM-IV regression output with interaction of bank liquidity and policy rate

Variable Regression #2

Zscore(-1) .2185795 (0.000)***

MPR(-1) -.5484064 (0.000)***
BSIZE(-1) -12.24055 (0.000)***
LIQ(-1) -18.15396 (0.000)***
ELNDGR(-1) .0172323 (0.000)***
PRFT(-1) 24.89666 (0.000)***
CAP(-1) -43.85155 (0.000)***
RGDP(-1) .926396 (0.000)***
MPR*LIQ -.7141919 (0.000)***

76
Wald Chi2 , Prob. 85417.83 (0.0000)***
Arellano-Bond Test for Order 1 -3.0612 (0.0022)***
Auto-
Order 2 -.76109 (0.4466)
correlation
Sargan’s Test 71.99342 (0.6400)
***, **, * = Significant at 1%, 5% and 10% respectively.The following diagnostic tests are run for the model: the Sargan test for
overidentification, the Arrelano-Bond test for autocorrelation and the Wald Chi2 for model fit. The results indicate that the over
identifying restrictions are valid, there is no first order or second order autocorrelation and the model fits the data well.

All the coefficient signs and the levels of significance in the basic model are maintained. The
coefficient of the interaction term between liquidity and the policy rate is negative and
significant. It appears that the negative impact of excess or very liquid positions on bank risk
behaviour is aggravated when interest rates begin to rise. In other words, when interest rates go
up, liquid banks are prone to more risky behaviour. From the study results, this may be an
indication that a significant portion of banks’ risky assets may be in longer term fixed rate
portfolios. Following from this premise then, as rates on the market rise, a funding and possible
maturity mismatch occurs: longer term assets may be yielding returns significantly below banks’
current liability costs, triggering a possible decline in profitability. An attempt to rebalance this
mismatch may cause the banks to take on default prone, but high yielding assets, thus causing
their risk profiles to deteriorate. Another perspective could be that in periods of rising interest
rates, the opportunity cost of idle cash may prove very high causing banks to invest in what
would have previously been considered risky projects just to maintain solvency, but which may
turn out to be counterproductive. On the whole it appears that risk-taking behaviour of banks
with significant liquidity is much more sensitive to monetary policy changes.

Table 4.4 presents the regression output of the main model augmented with an interaction term
of bank capital and the policy rate, MPR*CAP (Equation 4.3). The interaction term between
capital and the policy rate is positive and significant. It appears that in an environment of rising
interest rates, well-capitalized banks in SSA reduce their risk-taking behaviour, possibly a
reflection of how much capital they stand to lose when risky assets go bad. It seems that when
the stakes from possible loss, given default are high, banks with high equity capital engage in less
risky behaviour.

77
 Table 4.4: Regression output with interaction of bank capital and policy rate

Variable Regression#3
Coefficient p-value
Zscore(-1) .2059217 (0.000)***
MPR(-1) -.4855634 (0.000)***
BSIZE(-1) -5.99402 (0.000)***
LIQ(-1) -12.53074 (0.000)***
ELNDGR(-1) .0086003 (0.003)***
PRFT(-1) 8.356503 (0.204)
CAP(-1) -40.72834 (0.000)***
RGDP(-1) .8288938 (0.000)***
MPR*CAP 3.430164 (0.000)***
Wald Chi2, Prob. 171181.10 (0.0000)***
Arellano-Bond Order 1 -3.0347 (0.0024)***
Test for Auto- Order 2 -.88292 (0.3773)
correlation
Sargan’s Test 66.96571 (0.7859)
***, **, * = Significant at 1%, 5% and 10% respectively. The following diagnostic tests are run for the model: the Sargan test for over-
identification, the Arrelano-Bond test for autocorrelation and the Wald Chi2 for model fit. The results indicate that the over identifying
restrictions are valid, there is no first order or second order autocorrelation and the model fits the data well.

All other coefficients maintain their signs and significance in the model with capital and
monetary policy interaction terms, except for bank profitability which becomes insignificant in
the presence of the interaction term. The loss of significance of the bank’s profitability variable,
after including the interaction term could be an indication that in the presence of potentially
significant capital losses, the search for profits or yield in risky investment is tempered, hence the
loss of significance. In other words, capital and profits are not complements but rather
substitutes. Another conclusion that can be drawn, based on the results, is that well-capitalized
banks are less prone to risky behaviour as compared to their thinly capitalized counterparts.

The interaction term between bank size and the policy rate, MPR*BSIZE, in Table 4.5, has a
negative and significant coefficient.

Table 4.5: Regression output with interaction of bank size and policy rate

78
Variable Regression Coefficients p-value

Zscore(-1) .2351692 (0.000)***

MPR(-1) -.5138069 (0.000)***

BSIZE(-1) -9.439873 (0.000)***

LIQ(-1) -14.4197 (0.000)***

ELNDGR(-1) .0132871 (0.000)***

PRFT(-1) 21.92368 (0.001)***

CAP(-1) 42.03029 (0.000)***

RGDP(-1) .9442075 (0.000)***

MPR*BSIZE -.0290274 (0.002)***

Wald Chi2, Prob. 84868.02 (0.0000)***

Arellano-Bond Test for Order 1 -3.1539 (0.0016)***

Autocorrelation Order 2 -1.2762 (0.2019)

Sargan’s Test 69.91915 (0.7036)

***, **, * = Significant at 1%, 5% and 10% respectively. The following diagnostic tests are run for the model: the Sargan test for
overidentification, the Arrelano-Bond test for autocorrelation and the Wald Chi2 for model fit. The results indicate that the over
identifying restrictions are valid, there is no first order or second order autocorrelation and the model fits the data well.

Based on this finding, it appears that the risk positions of large banks also seem to deteriorate
when interest rates begin to rise. The possible reason for this occurrence could be that as interest
rates rise, the potential for profits also rises, albeit with higher risks. This potential for higher
profits may cause such large banks to try to compete for even larger shares of the market and
ultimately end up with more risk. This suggests that large banks in SSA are expected to engage in
more risky behaviour in an environment of rising interest rates and not the other way round. All
other coefficients maintain their signs and significance.

4.5 Conclusion and Policy Implications

From the findings presented so far, there is no evidence that a risk-taking channel of monetary
policy, as is posited in the literature, operates in SSA. It seems that the argument that periods of
low interest rates trigger risky bank behaviour may not necessarily hold for Sub-Saharan Africa.
We, however, argue that expansionary monetary policy of low interest rates may only constrain

79
risky behaviour in an environment of significantly low inflation. So though expansionary
monetary policy may present significant gains in the area of bank stability for SSA and
concurrently boost capital accumulation to spur economic growth, these findings should be
interpreted with caution since nominal and not real rates were modelled for this study. Bank-
level and macro-level characteristics also prove very important in explaining bank risk-taking
behaviour. Despite the absence of a risk-taking channel, as per the construct documented in the
literature, it can be concluded that monetary policy is not necessarily neutral with regards to the
financial stability of the SSA region. There is enough evidence to suggest that a change in the
monetary policy stance has implications for financial stability, especially in the banking sector.

In terms of policy implications, it appears that policy moves to reduce central bank rates may be
more beneficial to the region since it seems that raising interest rates in the SSA region could be
counter-productive and cause banks to engage in risky behaviour. Central banks therefore need
to preserve real rates of return when pursuing contractionary monetary policy of high interest
rates. The policy implications of rising interest rates need to be considered seriously given the
dominance of banks in the financial systems in SSA and the implications for the economy.
Improving the macro economy, i.e. increases in GDP, will also enhance the stability in the
financial sector. Thus, as much as possible, policy moves to improve the macroeconomic
environments should be pursued.

Another policy implication from the findings has to do with central bank supervision in periods
requiring contractionary monetary policy. In such periods, it is imperative that large banks, banks
with high liquidity and banks with small equity capital be monitored closely in order to curb
excessive risk taking behaviour. Controlling risky behaviour in large and liquid banks will help to
reduce potential bank failures and avert possible financial crises during periods of contractionary
monetary policy. It might also be appropriate to promulgate rules that control excessive growth
of unit banks within the sector. Increases in bank equity capital, whether by choice or as a result
of regulation, should be accompanied by closer monitoring so as to curb risky behaviour. To
conclude, it appears that the benefits of expansionary monetary policy may outweigh the
downsides where SSA is concerned, all else equal. This calls for a concerted move towards a
more expansionary monetary policy environment whenever feasible.

80
 CHAPTER FIVE
MONETARY POLICY, PRIVATE CAPITAL FORMATION AND ECONOMIC
GROWTH – THE CASE OF SUB-SAHARAN AFRICA

5.1 Introduction
Significant differences in the rates of growth in economies across the globe have, over the years,
stimulated quite a large volume of research into what drives economic growth. In the light of this
search, a number of economists have come up with a list of variables seeking to explain
economic growth. A few on the list include public infrastructure (Easterly and Rebelo, 1993;
Munnell, 1992; Aschauer, 1990), high returns to public infrastructure (Sanchez-Robles, 1998;
Kelly, 1997), educational attainment and the nature of political systems (Barro, 2002; Barro, 1999
and Barro and Lee, 1993), cultural factors (Harrison, 1992; Dieckmann, 1996), linguistic and
ethnic differences (Easterly and Levine, 1997), democracy (Barro, 2002; Barro, 1999), and even
religion (McCleary and Barro, 2006).

This preoccupation to find the drivers of economic growth also saw the emergence of numerous
theories seeking to explain the dynamics of economic growth. One of the prominent theories
that emerged is the endogenous growth model (Barro, 1990; Romer, 1990; Lucas 1988).
According to the endogenous growth theory, capital accumulation is key to economic growth.
Capital accumulation is in turn stimulated by the appropriate cost of capital. In following with
the cost of funds approach by de Bondt (2005), then the monetary policy rate, which is the
primary indicator for cost of funds in an economy, becomes one of the most important
determinants of its capital accumulation and ultimately its growth.

In spite the seemingly obvious consensus that monetary policy has a relevant role in
macroeconomics, Starr (2005), argues that the extent to which a country can use a monetary
policy to affect output is still an open empirical question. This is because even though the
available literature is seeking to provide a role for monetary policy in the growth debate, the
evidence, according to Fan et al., (2011), is unequally skewed in the direction of developed
economies, especially the US. It appears, therefore, that the role of monetary policy in the
growth dynamics of developing regions, like the Sub-Saharan African region, cannot be
necessarily taken as a given, and primarily so because of the dearth of supporting empirical
evidence. It is in this regard that this study seeks to empirically explore the question –‘does
monetary policy hold the key to capital accumulation and economic growth for SSA?’ In other

81
words, does monetary policy stimulate capital accumulation and ultimately economic growth in
the context of SSA, a region that has been lagging significantly behind the rest of the world in
terms of development?

Answers to these questions are obviously relevant given the view of Reed and Ghossoub (2012)
on the relevance of a tailored policy. They argue that if policy makers in the developing world
look to the experiences of advanced countries as a guide for policy making, then aggregate
activity in these countries will be even more distorted. This line of argument seems to suggest
that while aggregate activity in developing countries is already distorted, failure to devise tailored
policies to stimulate the needed growth will worsen the case. Zakir and Malik (2013) further
support this view by arguing that international experiences even provide reason to believe that
asymmetries are present with regard to the impact of monetary policy on the economy. These
asymmetries in response to aggregate demand, according to Morgan (1993), were evident in the
practical experiences of the US and Japan. In 1988 and 1989, for instance, a tight monetary
policy in the US succeeded in slowing down the economy, but an expansionary policy in 1990
failed to stimulate the economy. In Japan, in the late 1990s, an expansionary monetary policy also
failed to revive the economy that was in recession. In recent times, following the global financial
crisis, unusually low interest rates have not yet been able to trigger the required changes in
economic output in the US, for example. Zakir and Malik (2013) go on to intimate that the
asymmetries in response of aggregate output to changes in monetary policy can have strong
implications for an economy concerning issues such as the conduct of monetary policy and the
costs of changes in nominal demand. Such experiences obviously motivate a study that
empirically investigates the actual impact of monetary policy on economic performance for SSA.

This chapter sheds some light on some of these issues by looking for empirical evidence on the
real effects of monetary policy on economic output, and exploring its pathway through private
capital formation. The contributions of the study to the literature are thus three-fold: the first is
the findings on the impact of monetary policy on private capital formation in SSA, the second is
literature contribution to the debate on the link between private capital formation and economic
performance and the last is the real effects of monetary policy on economic output. The rest of
the chapter is structured in the following manner: the next section, section 2, addresses the
relevant literature. Section 3 provides the details of the study methodology. Sections 4 and 5 deal
with the discussions and policy implications respectively.

82
5.2 Review of the Literature
This section reviews the available theoretical and empirical literature on monetary policy of
interest rates, private capital formation and economic performance. The first section addresses
the relationship between interest rates and private capital formation, the second on private
capital formation and growth and the third on interest rates and economic growth.

5.2.1 Interest Rates and Private Investment

According to Ahmed and Islam (2006), economic theory suggests that changes in either nominal
or real interest rates exert a direct influence on investment spending. Monetary policy, according
to Kahn (1998), also works principally through its direct control of very short-term rates and
indirectly through long-term interest rates, the latter being the mode for affecting sensitive
sectors of the economy: housing, business investment and consumer durables among others.

On the issue of the posited relationship between interest rates and investment, though, there are
two very divergent views. There is the substitutability hypothesis, discussed in the works of
Tobin (1965) and Johnson (1967). They view money as a substitute for physical assets. Thus an
increase in the rates of interest on monetary assets would result in the shifting of portfolios
towards monetary assets with the consequences being a reduced rate of physical capital
accumulation. This view therefore holds that interest rates, being the primary user cost of capital,
must be low in order to stimulate fixed capital formation. In this regard as well, the thinking is
that monetary policy that facilitates affordable credit to the private sector will enhance capital
formation, while contractionary monetary policy, by making capital expensive, will reduce capital
formation. In essence, the relationship is negative.

The alternative view, however, places interest rates in relation to the savings (both domestic and
foreign) that provide the funds for investment. This view is predicated on the complementarity
hypothesis of McKinnon (1973) and Shaw (1973), which was later formalized by Galbis (1979)
and Fry (1980). According to this view, when real interest rates are realistic (positive), savings
mobilization is more effective and there is availability of loanable funds. The high cost of capital,
as a result of high nominal rates, also provides an inducement to efficiency in resource allocation
and ultimately higher productivity of capital. This is predicated on the thinking that because
investments take place when marginal productivity of the investment exceeds or equals the cost
of the investment, then a high cost of capital is associated with mostly highly productive
investments. Emery (1971) earlier advocated positive real rates of interest for developing

83
countries, given that these regions have a relative scarcity of funds in relation to their strong
demand for capital. He pointed out further, that low interest rates in these regions spurred
inefficiency and misallocation of economic resources. Following along these lines of reasoning,
the relationship between interest rates and capital formation is posited as being a positive one.

These differing perspectives on the responsiveness of investment spending to interest rate

movements, vis-à-vis the viewpoint of new growth theories that have investment spending as
key, have triggered a significant volume of research interest in search of supporting evidence.
The evidence, from a number of these studies, seem to indicate that aggregate level investment
spending in general is negatively related to interest rates. Shafik (1992) in a study on Egypt
arrives at the conclusion that a tight monetary policy is significantly and negatively related to
private capital formation. This finding was also corroborated by Ogunbenro et al., (1996) and by
Kumar and Mlambo (1995) for Nigeria and SSA, respectively. In a recent study for South Africa,
Khumalo (2014) documented a unidirectional causation running from interest rates to private
capital formation, and the relationship is negative and statistically significant. Despite the
seemingly overwhelming evidence of an inverse relationship between interest rates and capital
formation, some other studies such as Mehrara and Karsalari (2011) argue that this negative
effect is only evident when real rates exceed a threshold of 5 – 6%, otherwise the effect is
positive. They employed 101 developing countries for the period 1970 – 2010 and found the
relationship between real rates and investment to be that of an inverted U-form. Shapiro (1986)
and Baldawi (2005), on the other hand, actually concluded from their studies that aggregate level
investment spending is insensitive to interest rate movements. To further strengthen the point
with regard to the interest effect on capital accumulation, subsequent studies employed
disaggregated data and micro-level data. Upon disaggregating the investment data, some more
findings have emerged to support the negative link between investment and interest rates,
although most of these studies are skewed more towards developed economies. Chatelain et al.,
(2003) in a study on firm investment and monetary policy in the Euro area, for example,
concluded that monetary policy actually has negative and significant long-run effects on the
capital stock of firms in Germany, Italy, Belgium, Luxembourg, and France. Inferring from the
impact of user cost of capital, Akbar and Jamil (2012) also report that monetary policy has an
adverse but insignificant relationship with private investment in the agricultural sector in
Pakistan. In general, the view that seems to be supported by the available literature is the
substitutability hypothesis that advocates low interest rates to stimulate investment.

84
5.2.2 Private Investment and Economic Performance
In both developed and developing countries, private investment is considered one of the major
contributors to economic growth (Matwang’A Lusambili, 2000). In fact, according to Blomstrom
et al., (1996), economists and historians generally agree that there can be no fast long-run growth
without large investments in fixed capital. This is because increasing investment triggers the
adoption of new technology, creates new employment opportunities, grows incomes and, these,
ultimately lead to economic growth. In general, private investment is considered a very crucial
pre-condition for economic development because, by bringing together resources for the
production of goods and services, it allows economic activity to be set in motion. In light of
these, the question: ‘can a country create growth by enhancing fixed capital formation?’ has
received a lot of research interest. This interest is heightened more because of the notion that
private investment, when successfully harnessed, can play a very valuable role in reducing
poverty. A number of empirical studies support this view and further indicate that private sector-
led growth has an even more positive impact on growth compared to public investment. This
differential impact, according to Serven and Solimano (1990) and Coutinho and Gallo (1991), is
because public sector investment is relatively less efficient compared to private investment. In
response to the alleged efficiency of private investment in stimulating economic growth, a
number of studies have been conducted to ascertain the link.

De Long and Summers (1992) concluded from their study that the rate of capital formation
determines the rate of a country’s economic growth. In a study of East Asia, Young (1994)
concluded that investment was the main driver of the growth experienced in the East Asian
economies. Barro (1991), on the determinants of growth, also pointed out investment as one of
the many factors responsible. Gutiérrez (2005), in a study of the six largest countries in Latin
America also found that investment in machinery and equipment and private investment in
general were most effective in raising per capita GDP growth. M’Amanja and Morrissey (2006)
examined the determinants of growth for Kenya for the period 1964 – 2002. Using per capita
GDP to proxy growth, the study found that investment has a strong impact on growth.
Athukorala and Sen (2002) corroborate these findings, but add that the growth effect of private
investment is not only dependent on the volumes but also linked to the efficiency of the
investments. On the whole, these findings go to support the conventional argument that a
slowdown in investment translates into a slowdown in growth, suggesting a positive effect of
investment spending on economic performance.

85
5.2.3 Interest Rates and Economic Output
In terms of the relationship between interest rates and aggregate output, one of the conventional
economic theories has it that, as cost of capital becomes cheaper, investment spending increases
with an ultimate increase in growth. In essence, a contractionary monetary policy contracts the
economy while an expansionary monetary policy expands the economy via its effect on
investment. A second school of thought, however, holds that investment spending, which is the
foundation for growth, can only take place when there is funds availability. Funds availability is
in turn dependent on savings. According to this perspective, since savings can only take place in
an environment of positive real interest rates (high nominal rates in the presence of high
inflation), investment spending and growth are rather associated with high and not low nominal
interest rates. In other words, a contractionary monetary policy actually spurs savings, which lead
to investment and growth. Yet another view, which seems to be an extension of the second view,
is of the opinion that the relationship between interest rates and economic growth is more of a
non-linear one, with an inverted U shape. As per this perspective, at very low or very high levels
of interest rates, economic growth is hindered. They suggest that though growth and investment
can be positively associated with interest rates, these rates must operate within a threshold.

On the empirical level, the World Bank (1989) documented that the rate of economic growth in
countries with strongly negative real deposit rates is lower compared to countries with positive
real rates. In the same study, they documented higher productivity of investment in the higher
real rate environment (approximately 4 times more). According to this study, the potentially
negative effect of a low interest rate on growth is exerted, not only through the quantity of
investment, but also through the quality of investments. De Gregorio and Guidotti (1995) also
found the growth and interest rate relationship to resemble an inverted U-curve. This indicates
that low and negative real rates tend to cause financial disintermediation and a slowdown in
growth, following from the McKinnon and Shaw (1973) hypothesis. On the other hand, when
very high real interest rates are not a reflection improved investment efficiency but point more to
a lack of credibility with regards to economic policy or other forms of country risks, there is a
likelihood that growth may be hampered. Fry (1997a) confirms the inverted U-curve
relationship, using a sample of 85 developing countries over the period 1970 – 1995. He
concludes that growth is maximised when real rates lie within the range of -5% to +15%. Shafik
and Jalali (1991) empirically corroborate this positive relationship between interest rates and
growth. They explain the positive relationship as being indicative of improved investment
efficiency even when investment volumes decline.

86
In line with the conventional argument on cost of capital effect, Van Els et al., (2001) found that
after an increase in the policy-controlled interest rate in the euro area, the decrease in investment
accounted for between 30 to 50 percent of the drop in GDP during the first year. This share
rises to as much as 80 percent of the fall in GDP after three years. In essence, a rise in the policy
interest rate causes a reduction in growth. Using data spanning 111 years (1901 – 2011), Hansen
and Seshadri (2013) also find a negative correlation between interest rates and GDP growth.
Along the same lines, Nouri and Samimi (2011) find a positive and significant relationship
between expansionary monetary policy and growth.

Looking at the empirical evidence so far, there is no doubt that the relationship between
monetary policy and economic performance is still ambiguous and the debate is still ongoing, yet
to arrive at a consensus.

5.3 Methodology
This section presents the details of the methodology employed for the study. The first part deals
with the model and estimation approach, followed by the empirical justification of the variables
included in the model.

5.3.1 Modelling and Estimation Approach

In order to estimate the effect of monetary policy on economic performance and private capital
formation in SSA, nine countries are sampled. The sample is based purely on the availability of
data for the period under study: 1999 – 2011. The countries are: Botswana, Ghana, Kenya,
Mauritius, Namibia, Nigeria, South Africa, Swaziland and Zambia. The variables for the study are
GDP per capita, private capital formation (as a percentage of GDP), monetary policy rate
(annual average), stock market capitalization (as a percentage of GDP) and openness (exports
plus imports as a percentage of GDP). The series were drawn from the World Development
Indicators (WDI) and the International Financial Statistics (IFS).

In terms of model selection, the study adopted the growth model by Zakir and Malik (2013). In
their model, output is a function of its own lag, monetary policy and the standard determinants
of growth: terms of trade, stock market development, private investment (Khan and Senhadji,
2000). The private capital formation model for the study follows the adaptation by Adjasi and
Biekpe (2009) of the traditional simple investment model by Barro (1990). Following the user

87
cost of capital and investment literature (Mishkin, 2007; Ribeiro and Teixeira, 2001; Iacoviello,
2000 and others), this investment model is augmented with the monetary policy variable, mainly
to capture the effects of interest rate on investment. The study thus arrives at the following
models for growth and capital formation respectively:

GDPCit  0  1MPRit   2 Pr ivcapit  3Mktcapit   4Tradeit  5GDPCit 1  et (5.1)

Pr ivCapit  0  1MPRit   2 Mktcapit  3Tradeit  5 Pr ivCapit 1  et (5.2)

To explore the effect of monetary policy on growth via private capital formation, the monetary
policy variable is interacted with private capital formation in model 1 to arrive at:

GDPCit   0  1 MPRit   2 Pr ivcapit   3 MPR * Pr ivcapit   4 Mktcapit   5Tradeit   6 GDPCit 1  et

………….
(5.3)

Where GDPCit is GDP per capita for country i in time t. MPRit is the monetary policy rate to proxy for
the stance of monetary policy, Privcapit is private capital scaled by GDP, Mktcapit is stock market
capitalization scaled by GDP, Tradeit is imports plus exports scaled by GDP and MPR*Privcapit is the
interaction term between private capital and the monetary policy rate. GDPCit-1is the lag of GDP per
capita in time t for country i.

The equations are estimated as dynamic panels, where a lag of the dependent variable is also
included as an independent variable. This dynamic panel setting creates endogeneity issues. To
address the potential issues of endogeneity and possible omitted variable bias that may arise from
the model, the estimation is done using the more efficient system generalized method of
moments (GMM) by Blundell and Bond (1998). This simultaneously corrects for endogeneity
issues that arise because of a two-way effect among the variables of study, and unobserved
heterogeneity issues. The system GMM uses lagged values of the regressors (in levels and in
differenced form) as instruments for right-hand side variables and removes the fixed effect from
the error term.

88
5.3.2 Justification of Variables
5.3.2.1 Central Bank Policy Rate
Based on the theoretical literature, the relationship between interest rates and investment and
economic growth can be either negative or positive. In the former case, which follows the Barro-
Becker path, high interest rates and ultimately high cost of capital is a deterrent to investment
and growth. In the latter case, the relationship to investment and growth works via savings. In
this instance, therefore, high interest rates attract savings which serve as loanable funds for
investment and ultimately promote growth. Like the theoretical literature, the empirical evidence
on interest rates, private capital formation and growth is also yet to arrive at a consensus. While
some studies have arrived at a strongly negative relationship between interest rates on capital
formation and growth (Khumalo, 2014; Hansen and Seshadri, 2013; Van Els et al., 2001) and
some others have documented a positive relationship (Fry, 1988; Asante, 2000; Munir et al., 2010;
World Bank, 1989; De Gregorio and Guidotti, 1991), yet some have documented a non-linear
relationship (Mehrara and Karsalari, 2011; Fry, 1997a). Based on the evidence so far, there is no
a priori expected sign for the impact of monetary policy on investment and growth.

5.3.2.2 Stock Market Development

Stock markets, according to Adjasi and Biekpe (2009), enable firms to quickly and efficiently
acquire capital which is channelled into profitable projects, thus making a significant
contribution to economic performance. Tobin (1969) and Atje and Jovanic (1993) also point to a
positive link between stock market development and investment. In contradiction to the above,
there are theoretical arguments, that the services provided by the stock markets can also hinder
economic growth (Demirgüç-Kunt and Levine, 1996). Devereux and Smith (1994), for example,
believe it reduces saving rates through income and substitution effects, thus hindering economic
growth. Another view also holds that economic development is thwarted when dissatisfied
participants find it easy and quick to sell in a liquid equity market.

Empirical investigations into the link between stock market development and investment and
economic growth also seem to be far from conclusive. While Yartey (2008) and Rousseau and
Wachtel (2000) find that stock market development is positively influential on investment and
growth, Stiglitz (1994), Shleifer and Vishny (1986) and Bhide (1993) find otherwise. Adjasi and
Biekpe (2006) find mostly insignificant relationships for their growth models and significant
relationships in their investment models. The coefficient sign in this study is thus left open.

89
5.3.2.3 Private Capital Formation
The link between private capital formation and economic growth is a relatively well-documented
one. The theory holds that private capital is more efficient compared to public capital and spurs
higher productivity and ultimately growth. According to Athukorala and Sen (2002) this
productivity effect is however dependent not only on the volume of investment but also on (the)
efficiency. There is a significant volume of empirical studies to support the conventional
argument that private capital formation fosters growth (Blomstrom et al., 1996; De Long and
Summers 1992; Young, 1994; Gutiérrez 2005; M’Amanja and Morrissey, 2006). On the basis of
the empirical evidence, the a priori expectation of private investment on growth is positive.

5.3.2.4 Trade
With the emergence of the endogenous growth theories, the role of government policy in
stimulating growth has received significant research attention. The proponents of liberalization
policies, for example, advocate trade openness on account of its benefits: reduction in
production inefficiencies (Roubini and Sala-I-Martin, 1991; Kar et al., 2008); increased
competition in the domestic economy and hence productivity, and enlarged market for domestic
producers to enjoy scale economies (Taylor, 1994). These benefits are based on the import
discipline hypothesis (Aghion, Harris and Vickers, 1997). Serven (2002) is, however, of the view
that “an abrupt increase in exposure to external competition in certain sectors can make these
sectors less attractive as a destination for new capital flows” and ultimately be inimical to growth.
Also, where openness to trade is accompanied by currency depreciations, then capital formation,
that is reliant on imports, will suffer and negatively affect growth. On account of the arguments
above, the study does not set an a priori expectation on the effect of trade openness on private
capital formation and growth.

5.4 Findings and Discussion

This section presents the analysis and discussion of study findings. It also deals with the
summary statistics of the variables under study, the correlation matrix of the variables and finally
the regression output for the three equations.

90
5.4.1 Descriptives and Correlation Analysis
In terms of the descriptives, Table 5.1 indicates that the average GDP per capita for the nine4
countries that were studied averaged approximately 2.69 over the 13-year period. Some countries
were, however, as low as -8.690 and others as high as 30.344. For private capital formation, the
highest value over the study period was 20.66 percent of GDP and a low of 4.06% of GDP.
The period average was 12.61% of GDP. It appears that investment spending in SSA is not
significant in relation to GDP.

Table 5.1: Summary Statistics of Study Variables (1999-2011)

Variable GDP_per Privcap MPR MktCap Trade

capita
Mean 2.698681 12.61943 12.87535 39.75085 89.65551
Std dev. 3.750647 4.469236 6.367084 59.43053 33.67453
Min. -8.690764 4.066759 2.96 4.243549 42.65139
Max. 30.34408 20.66257 40.0981 291.2753 202.8499
Source: author computations.

The proxy for the monetary policy stance, the central bank policy rate, had a mean of 12.87%
over the period, with minimum and maximum values of 2.96% and 40.09% respectively. Stock
market development, proxied by market capitalization over GDP, was approximately 39.75% on
average for the sample. Some countries, however, had low ratios, evidenced in the minimum
value of 4.24% of GDP while others exhibited significant stock market investment (291.2% of
GDP) over the same period. Like stock market development, trade openness also exhibited
some significant differences among the countries under study. For instance, in spite of a
maximum value of 202.84% of GDP within the study sample, indicating a very open economy,
the minimum value stood at 42.65% of GDP, roughly one-fifth of the maximum value. The
average trade share value for the sample for the period stood at 89.6555% of GDP.

To test for the possibility of multicollinearity in the regressors, the correlation between the
variables were tested (see Table 5.2). From Table 5.2, private fixed capital formation exhibits a
positive correlation with GDP per capita and trade openness, but a negative relation with the
central bank policy rate and stock market capitalization.

4
Botswana; Ghana; Kenya; Mauritius; Namibia; Nigeria; South Africa; Swaziland; Zambia

91
Table 5.2: Correlation Matrix of Study Variables

Variable GDP_percapita PrivCap Mpr MktCap Trade

GDP_percapita 1.0000
PrivCap 0.1512 1.0000
Mpr 0.0049 -0.0401 1.0000
MktCap -0.0360 -0.2302** -0.2030** 1.0000
Trade -0.1059 0.1522 0.0653 -0.3098*** 1.0000
*, **, *** = significance at 10%, 5%, and 1% respectively

Of these relationships with private capital formation, only the relationship with stock market
capitalization was significant at 5%. The monetary policy rate has a positive and insignificant
correlation with GDP per capita and trade openness but a significant and negative correlation
with stock market development. Trade openness further exhibits negative correlations with
GDP per capita and stock market development, the latter relationship being significant at 1%;
the relationship with the policy rate is, however, positive but insignificant. None of the
correlations are high enough to raise concerns of multicollinearity in the regression.

5.4.2 Regression Analysis

The results of the dynamic panel data regression of monetary policy on economic performance
reveal some interesting relationships. There is a negatively significant relationship between the
lag of GDP per capita and its current value. The significance points to the fact that past levels of
GDP per capita help to explain current levels while the negative sign is indicative of cyclicality.
This seems to suggest that periods of growth and contractions seem to alternate in SSA or that a
period of increased growth is immediately followed by a period characterised by decreased
growth. In other words, GDP growth is not a smooth or persistent occurrence. According to
Lipsey and Chrystal (2011), such cyclicality results from deviations in the trend of growth in
GDP and shifts in the aggregate demand curve.

The monetary policy variable, from Table 5.2, has a positive and significant coefficient. In terms
of interpretation, it appears that increasing interest rates and thus higher nominal cost of capital
actually stimulates growth. This finding is obviously contrary to the more general literature
(Hansen and Seshadri, 2013; Nouri and Samimi, 2011; Van Els, 2001, and Barro, 1991) that has

92
mostly documented a strongly negative relationship, or at best an insignificant relationship, but
still with a negative sign. By inference from the sign and significance, however, it may appear that
while the nominal interest rates rise, real interest rates may still be low or negative, given that
most of these economies also experience high inflation rates. In such instances, the high nominal
rates may not necessarily stifle productivity growth, as the real cost of capital still stays low.
Contrary as the study finding may be to the conventional argument, it supports the findings of
studies by the World Bank (1989), De Gregorio and Guidotti, (1995) and Fry (1997a) that
interest rates and growth are positively related. The explanation of the positive link between
growth and interest rates was indicated as flowing more from the high marginal productivity of
investment capital and not just from the volume of the investment that feeds growth. In this
instance, while the volume may fall because of capital costs, productivity of the acquired capital
exceeds the drop in capital formation and the net effect on growth remains positive.

The other intuition for the positive effect of contractionary monetary policy on economic
performance, according to Emery (1971), flows more from the ability of high interest rates to
attract both domestic and foreign savings to fund growth. In other words, when interest rates are
high, savings increase and provide the needed funding for investment and thus growth. The
reasoning follows further, that high interest rates hold the potential to spur more labour-
intensive productivity as opposed to capital-intensive productivity, thus creating employment
and stimulating consumer spending, the most significant component of aggregate demand.
Emery (1971) offers yet another explanation for this puzzling evidence. He argues that low
interest rates stimulate inefficient capital accumulation, which may not necessarily foster growth.
He also suggests that when economic agents fail to pay a realistic cost of capital because of low
interest rates, their resource allocation is less than efficient. Following from Emery (1971) and
the other arguments above, the positive link can be attributed to more efficient savings
mobilization coupled with higher productivity of both labour and fixed capital in the SSA region.
The positive sign does not, however, rule out the possibility of a non-linear relationship between
interest rates and growth.

The proxy for stock market development, market capitalization to GDP, returns a statistically
insignificant and negative relationship with economic growth (Table 5.3). From the results, it
appears that developing the stock market is inconsequential in promoting the economic growth
of SSA. While this contradicts the empirical studies that support the finance-leads growth
hypothesis (Filer et al., 1999; Yartey, 2008), it does support the findings of Singh (1997) and

93
Levine and Zervos (1996), that financial development, like development on the stock market, has
little or even potentially negative effects on economic growth. Hoti and Nuhiu(2011) also argue
that the negative association with the stock market and growth may flow from unfavourable
conditions associated with the stock market, such as price volatility, illiquidity, inappropriate
regulation or organization. Bawumia et al., (2008) and Senbet and Otchere (2008) further concur
by indicating that the negative effect is due to the underdevelopment of these stock markets,
hence serving as a drain instead of a driver of economic growth. Another explanation can be
Azarmi et al.’s (2005) notion that the relevance of the stock market for economic growth is a
function of the policies prevalent at the time of study. In essence, if the policies do not
complement the stock market, the growth effect of the stock market will be negative. These
arguments ultimately raise the question on the role of effective institutions in the growth process.

Table 5.3: Regression output for growth in SSA

Regression #1
Dependent Variable GDP_PerCapita
Coefficient p-value
GDP (-1) -.1817796 (0.036)**
Mpr .4345404 (0.000)***
Mktcap -.0307553 (0.211)
PrivCap .3929374 (0.043)**
Trade -.058466 (0.145)
Wald Chi, Prob. 56.34 (0.0003)***

Sargan’s Test 78.92551 (0.0000)***

***, **, * = Significant at 1%, 5% and 10% respectively. GDP is GDP per capita; MPR is monetary policy rate;
MKTCAP is stock market capitalization; PrivCap is private capital formation; Trade is international trade. The
following diagnostic tests are run for the model: the Sargan test for overidentification, the Arrelano-Bond test for
autocorrelation and the Wald Chi2 for model fit. The results indicate that the over identifying restrictions are valid,
there is no first order or second order autocorrelation and the model fits the data well.

Private fixed capital formation, on the other hand, shows the expected positive sign and is
significant. By inference, increasing levels of fixed capital formation in the private sector
stimulates aggregate demand and vice versa. This finding is in consonance with Adjasi and
Biekpe (2009), Gutiérrez (2005), Barro (1991), M’Amanja and Morrissey (2006). In essence, the
finding seems to indicate that there can be no significant growth without investments in fixed

94
and productive capital or as De Long and Summers (1992) put it, capital formation determines
the rate of a country’s economic growth. It is therefore expected that as more and more private
capital is accumulated in productive areas, economic growth rates will be enhanced.

Trade openness has a negative and insignificant coefficient, contrary to the expectations of the
liberalization schools of thought (Barro, 1991). By inference, increasingly becoming open to
global trade does not hold any benefits for SSA. This could very well be a reflection of the very
uncompetitive position of SSA with regards to global trading. By being mostly net importers and
with exports of mainly primary products, trade openness allows for goods that compete with
growth stimulating sectors of the economy to flood the market and ultimately crowd out local
industries, suppress employment and impact growth negatively. The insignificance of the trade
variable also confirms the findings of Adjasi and Biekpe (2009) for SSA. So while it is not critical
that the SSA region be open to freer trade, the effects could be more of a negative than positive
effect for growth. The findings may, however, be in support of Baliamoune-Lutz and
Ndikumana (2007) that the limited effect of trade on growth is due to the weakness of relevant
institutions. They also hold the view that trade and growth have a U-shaped relationship. So at
the pre-threshold levels of trade, the effect of increasing trade is negative, and it is only after a
certain magnitude of trade that growth effects become positive. Mazumdar’s (1996) stance builds
on the magnitude of trade argument by linking the growth effects of trade to the composition of
trade, in other words, the quality of the trade basket. It may therefore be concluded, given that
the countries under study have mostly primary goods in their trade basket, that the features that
characterize the whole trade scenario contribute to its negative effect on growth.

The second model, (Regression # 2, Table 5.4) for this study, regresses private capital formation
on its own lag, the policy rate, stock market development and trade openness.

Table 5.4: Regression Results for capital formation for SSA

Regression #2
Variable PrivCap
PrivCap (-1) .8054611 (0.000)***
Mpr -.009169 (0.897)
MktCap -.0230604 (0.100)
Trade .0348031 (0.072)*

95
 Wald Chi, Prob. 520.70 (0.0000)***
Sargan’s Test 43.40852 (0.0028)***
***, **, * = Significant at 1%, 5% and 10% respectively. PrivCap is private capital formation; MPR is monetary
policy rate; MKTCAP is stock market capitalization; Trade is international trade. The following diagnostic tests are
run for the model: the Sargan test for overidentification, the Arrelano-Bond test for autocorrelation and the Wald
Chi2 for model fit. The results indicate that the over identifying restrictions are valid, there is no first order or
second order autocorrelation and the model fits the data well.

From Table 5.4, the coefficient of the lag of the dependent variable is positive and significant. It
may be indicative that private capital formation in SSA is a persistent phenomenon. This finding
conforms to the findings of Adjasi and Biekpe (2009) that current levels of private capital
formation are positively influenced by its past levels. It also suggests that there is year on year
improvement in fixed capital formation in the private sector and this function is very smooth. It
appears that SSA is steadily and persistently accumulating private capital, and this is fundamental
to long-term growth. It also appears that bursts of investment in the previous period spill over to
the subsequent periods. The policy rate, the proxy for the cost of access to capital in the SSA
regions, shows the expected negative sign albeit without any of the appropriate significance
levels. This implies that the policy rate does not in any way influence capital formation in the
private sector. It may also be indicative of the fact that while a contractionary monetary policy
could be inimical to private capital formation, it is obviously not one of the critical factors that
drive capital formation, hence the statistical insignificance. This finding supports the findings of
Shapiro (1986) that aggregate level investment is insensitive to interest rate movements.
According to Gelb (1989), such a relationship could emerge if capital formation is conditioned
on other factors like the productivity of the investment, the investment need and not just the
cost of capital.

Stock market development, from Table 5.4, is insignificant in relation to private capital
formation. Given the findings, there is the possible indication that funds that find their way on
to the stock market do not flow into the private sector of the region and if they do, they do not
end in fixed capital formation. It may also suggest that the stock markets compete for the
available investible funds and crowd out private sector capital formation. This finding contradicts
the findings of Adjasi and Biekpe (2009) on the positive association between stock market
development and private capital formation.

96
Trade openness is positive and significant in this regression. By interpretation, the increasing
share of foreign trade to GDP enhances private capital formation. This could be indicative of the
fact that as the trade openness enlarges the market for domestic producers, they respond with an
increase in capital accumulation so as to enjoy scale economies (Taylor, 1994). It could also be
that the reduction in trade barriers through liberalization creates an advantage to the export
sector and thus increases the incentive to accumulate capital.

The third model, (see Table 5.5), which is model 1 augmented with an interaction term between
monetary policy and private capital formation, presents similar results. The regression output
does not show any sign changes, though one of the insignificant variables in model 1 became
significant in model 3 and that is the trade variable.

Table 5.5: Regression output for growth in SSA

Regression #3
Variable GDP_PerCapita
GDP (-1) -.2063823 (0.018)**
Mpr .7940917 (0.000)***
MktCap -.0331377 (0.175)
PrivCap .6208867 (0.011)**
Trade -.0795867 (0.058)*
PrivCap*Mpr -.0231086 (0.133)
Wald Chi, Prob. 59.66 (0.0000)***
Sargan’s Test 78.02389 (0.0000)***
***, **, * = Significant at 1%, 5% and 10% respectively. GDP is GDP per capita; MPR is monetary policy rate;
MKTCAP is stock market capitalization; PrivCap is private capital formation; Trade is international trade.
PrivCap*Mpr is the interaction term between monetary policy and private capital formation. The following
diagnostic tests are run for the model: the Sargan test for overidentification, the Arrelano-Bond test for
autocorrelation and the Wald Chi2 for model fit. The results indicate that the over identifying restrictions are valid,
there is no first order or second order autocorrelation and the model fits the data well.

Given the negative and significant trade coefficient, it appears that for SSA, as the economies
become more open to foreign trade, growth in the region reduces. This confirms the view of
Serven (2002) on the possibility that foreign competition in certain sectors can make these
sectors less attractive as a destination for new capital flows and ultimately be inimical to growth.

97
Also, it is possible that because currency depreciation is common to the sample, imports for
production that stimulates growth is stifled due to high import prices. Another inference could
be the stifling of local industries and growth due to increased foreign competition.

The interaction term between private capital and the policy rate is negatively signed but
statistically insignificant, thus confirming what appears to be an inconsequential relationship
between private capital formation and monetary policy in SSA. The statistical insignificance,
notwithstanding increasing policy rates, can potentially serve as a tax on private capital
accumulation and compromise its role in growth. Given the potentially negative impact of the
interaction term, it can also be adduced that the economic growth prospects of less capital
endowed economies are quite sensitive to monetary policy hikes. It may also suggest, that though
private capital is likely to be negatively affected by high cost of capital, the impact is not
significant enough to pass-through to growth and also possibly because other factor may be
more relevant in driving the kind of capital accumulation that fosters growth and not necessarily
monetary policy actions.

5.5 Summary, Conclusion and Policy Implications

The study sought to test the impact of the policy rate on growth and investment in some selected
countries in the SSA region. Data for nine (9) countries over a 13-year period was captured and a
dynamic panel model estimated. Based on the study results, the major findings were: growth in
SSA is cyclical whilst private capital formation is a persistent and smooth function in the region.
Growth is positively related to the policy rate, following the McKinnon and Shaw (1973)
hypothesis, but stock market development has no effect on growth or private capital formation.
The overall impact of trade openness is ambiguous and the policy rate is not a critical factor to
private capital formation in the SSA region.

Based on the purpose of the study it can be concluded, that while private capital is obviously
positively influential on growth, the formation of private capital to stimulate growth is not driven
by the monetary policy stance.

In terms of policy implications, the findings seem to advocate high policy rates. While the high
policy rate may hold negative implications for private capital formation, the highly positive
relationship with growth may be indicative of its potential to attract both foreign and domestic
savings to fund growth. Such realistic market rates will also stimulate a better and more efficient

98
utilization of acquired capital. It will also stimulate higher employment of the labour force in the
region and thus reduce unemployment and stimulate consumption and growth. Empirical testing
for the point, when the potentially negative effect of the policy rate on investments turns
significant, may be relevant in identifying the threshold for target policy rates in the region.

Another policy implication flowing from the positive link between private capital formation and
growth is the need to stimulate private capital formation as it leads to growth. Policies to
encourage capital formation should be aggressively pursued. On account of the insignificant
relationship between capital formation and the policy rate, there may be the need to further
investigate the other determinants that stimulate private capital formation. Stock markets in the
SSA region seem to be failing in their functions because they are mostly undercapitalized, thinly
traded and illiquid. Policy moves to address these structural failures may work to reverse the
potentially negative effect found in this study.

On the issue of trade and the ambiguity of the findings, the argument is that while there may be
favourable effects of trade on aggregate capital accumulation, trade may not necessarily ensure
capital accumulation in sectors that are growth-enhancing hence the negative effect. The
ambiguity also calls for selective openness. In other words, sectors that can be harmed when
opened to foreign trade should be protected, whilst those that will benefit from being opened to
free trade can be liberalized. In essence, openness should be tailored as per target sector. This
kind of policy will require further study to identify sectors that will or will not benefit from
openness and be managed as such.

99
 CHAPTER SIX
SUMMARY, CONCLUSION AND RECOMMENDATIONS

6.1 Introduction
This final chapter summarises the important findings and policy implications emerging from the
various chapters constituting the dissertation. The conclusions that are drawn are based on the
findings of the study. In line with the broad objectives of the dissertation, relevant policy
recommendations are suggested, along with recommendations for future research.

6.2 Summary of Purpose and Findings

This dissertation is a collection of four stand-alone empirical essays exploring the monetary
transmission mechanism in SSA. The dissertation examines both micro level and macro level
responses to changes in monetary policy. Chapter One provides the overview while chapters two
to five are the empirical studies examining some issues on the monetary transmission
mechanism: the interest rate pass-through, exchange market pressure, bank risk-taking
behaviour, and the response of private capital formation and economic growth to changes in the
policy rate.

Chapter Two assessed the fullness of the interest rate transmission mechanism of the monetary
transmission mechanism. The chapter examined the effectiveness of the monetary policy
transmission to retail rates as being a function of the level of financial development. This
particular paper was limited to Anglophone West Africa, mainly because these countries seek to
integrate their monetary frameworks. The summary statistics revealed differences in the levels of
financial development, the policy rates and economic growth. These differences ultimately came
to bear on the study results. The study revealed significant differences in the transmission
mechanisms of the 3 countries that were studied: Gambia, Ghana and Nigeria. Whilst Ghana and
Gambia were mostly characterised by sluggish responses in the lending rate to policy moves,
Nigeria was characterised by overshooting in its lending rate response to policy changes. The
findings indicate a pass-through of less than unity for Gambia and Ghana but greater than unity
for Nigeria. In general the study failed to find an effective interest-rate pass-through in
Anglophone West Africa. That aside, there were also significant differences in the effect of the
level of financial development on the pass-through, thus the study failed to provide concrete
evidence to support the argument that the effectiveness of pass-through is dependent on the
level of financial development.

100
Chapter Three examines exchange market pressure on a currency as a function of the policy rate
and some macroeconomic fundamentals. Using a panel dataset on frontier markets for which
data was available, the paper sought to test the hypothesis that depreciation pressure on domestic
currencies in SSA can be eased with monetary policy changes. The regression results showed the
exchange market pressure on a currency as being dependent on its own lag, the policy rate, the
terms of trade, the stock of public and the current account balance. The results revealed that a
contractionary monetary policy can stave off depreciation pressure. The results also indicate that
GDP growth in SSA exerts depreciation pressure as far as domestic currencies are concerned.
Real GDP growth was significantly and statistically positively related to exchange market
pressure. The results also showed an appreciative effect of the stock of debt, the terms of trade
and the current account balance on domestic currencies. Private capital flows were insignificant
across all the regressions, including the ones that excluded South Africa, to test for the possibility
of biases in the estimates. The results of the chapter seem to support the notion that higher
domestic interest rates relative to other countries attract foreign capital in search of high
investment returns and subsequently cause domestic currencies to appreciate. The findings also
point to the fact that macroeconomic fundamentals are important in explaining currency
position in the foreign exchange market.

Chapter Four investigates the risk-taking behaviour of banks as a function of the monetary
policy rate, real GDP growth and bank characteristics, like liquidity, capital, lending and size. The
chapter sought to test for the presence of a risk-taking channel of monetary policy transmission
in SSA. The study also sought to test for the possibility that certain bank-level characteristics
predispose banks to risky behaviour. The results of the study failed to support the risk-taking
channel in the manner that is hypothesized in the literature: ‘that low interest rates trigger a
search for higher yield and make banks risky’. Instead, the results show that banks engage in
risky behaviour when interest rates begin to rise. This evidence is, however, unequivocal as the
behaviour of the banks in response to the contractionary policy environment may still be a
function of real interest rates. The results showed that a bank’s risk profile is a function that is
significantly dependent on its history, the size of the bank, the bank’s liquidity, its lending growth
and its economic growth. The results also suggest that real GDP growth, lending growth and
profitability are positively and significantly associated with improved risk profiles of banks, while
bank size, bank liquidity and bank capital drive risky behaviour. The use of interaction terms also

101
revealed that the large banks, and banks with high liquidity are more prone to risky behaviour
when interest rates are high.

Chapter Five examined three interrelated issues. The first part investigated the relationship
between the monetary policy rate and economic growth; the second, the relationship between
the policy rate and private capital formation; and the last part, the link between private
investment and growth, given monetary policy changes. In the first case, the results show that
economic growth is a function of its own lag, the policy rate and the level of private capital
formation. GDP per capita was negatively and significantly related to its lag, but positively and
significantly related to the policy rate and private capital. In the second case, private capital
formation emerged as a function of itself and the share of foreign trade. In the third case, growth
emerged as being influenced by its own lag, the policy rate, private capital formation and the
share of foreign trade. In all the two growth regressions, the lag of GDP per capita was
significantly negative, indicating cyclicality. Private capital formation was positively influenced by
its lag, indicating investment persistence. The policy rate emerged as positively significant in all
the growth regressions, failing to support the hypothesis that at high interest rates, growth
diminishes. The share of foreign trade shows up as positively significant in explaining investment
but negatively significant in explaining economic growth. The study fails to find a growth impact
for stock market development as well as confirm private capital as a function of interest rates.

6.3 Recommendations
Based on the findings made from the study, the following recommendations are suggested with
the aim that it will enhance the effectiveness of monetary policy in the pursuit of economic
growth. It is important that central banks find the appropriate trade-off between conflicting
goals that require changes to the policy rate. This will ensure a policy design that has more
benefits than costs. A case in point, for instance, is with some of the findings of this study: while
increasing the policy rate may be beneficial for economic growth and the domestic currency
position, it has the potential of increasing bank risk-taking behaviour and of possibly lowering
private investment. Given that bank risk-taking behaviour can trigger a financial crisis, reduce
private investments and retard growth, there is the need to assess the costs and benefits of each
policy move before it is implemented. In this particular regards, it might be prudent for central
banks to ensure that real rates of return are preserved in periods of rising policy rates. This is to
minimise the potential risk-taking effect that nominal rate hikes may trigger.

102
It is also important for Sub-Saharan Africa to pursue policies that will improve upon their trade
positions. Currently, most of the countries are net importers. Policies that will transform them to
net exporters might prove beneficial in the long run. Steps to add more value to the current
tradables in SSA might also prove valuable. On the same issue of trade, it might be imperative
for Sub-Sahara Africa to re-evaluate its trade openness with the view to protect some sensitive
sectors of the economies. This will ensure, that whilst still being open to trade in the globalized
world, the free trade agreements are tailored to bring the best of benefits to the table for Sub-
Saharan Africa.

In a related framework, it might also prove beneficial to encourage consumption of locally

manufactured goods. This can either be through restrictions on imported substitutes, or if price
is the determining factor, then targeted subsidies to make the local alternative competitive. This
will ensure that growth is accompanied by strong domestic currencies. Policies should also be
directed at improving the export capacity of the Sub-Sahara African region, both in terms of
volumes and value. This will also lend strength to the currencies.

Furthermore, it should be noted that monetary policy is not neutral in terms of financial stability.
In periods of rising interest rates, when banks are predisposed to potentially risky behaviour,
large banks, liquid bank and under-capitalized banks need to be monitored even more closely to
control overly risky behaviour. Banks with these characteristics are also obviously very important
for the stability of banking systems, given that size, liquidity and adequate capital are the pillars
of stable financial systems. As such, prudent supervision during contractionary monetary period
should be appropriately designed to ensure that the stability of the financial system is not in
jeopardy.

The test of the interest rate pass-through, which is one of the earlier points of the monetary
transmission mechanism (MTM), reveals a lot of deficiencies in West Africa. It might be
beneficial for the West African sub-region to improve on or establish, where non-existent, the
channels that facilitate the interest rate pass-through: stock markets and bond markets. Credit to
the private sector, which also serves as a conduit for the policy rate transmission and is the
measure of financial development in this study, should be improved upon. In essence, as more
credit flows through to the private sector, a medium for policy transmission is created.

103
Private capital formation should be encouraged through policy moves to protect property rights
and ensure appropriate legal recourse where needed. In essence, institutional weaknesses such as
contractual incompleteness, corruption, lack of well-defined property rights and poor
enforcement of contracts should be corrected.
A reasonable certainty on good policies being implemented could also work to create the
appropriate climate for investment planning.

6.4 Conclusions and Recommendations for Future Research

On the whole, this dissertation makes a case for the relevance of monetary policy in the pursuit
of a macroeconomic agenda as well as in the behaviour of economic agents. A key implication is
that monetary policy is not neutral in the pursuit of economic goals, both on the macro level and
on the micro level. The collective thrust of this dissertation is the confirmation of the potentially
conflicting effects of a particular stance of monetary policy. This raises the ultimate question,
how do we find the optimal trade-offs with regards to the effects of monetary policy? This could
form the foundation for future research.

A limitation in this dissertation that could provide direction for further research is in the choice
of the variable to proxy monetary policy. While this measure, the policy rate, may be adequate
within the context of the study objectives, a major concern is the possibility of losing valuable
information on the transmission mechanism because other measures of monetary policy, such as
the money supply or real interest rates, for example, were not employed. In this regard, future
research could still explore the areas within the study but this time employing other measures of
monetary policy. Another area worth exploring is the transmission mechanism as it operates in
the rather large informal financial sectors across Sub-Saharan Africa. This may offer better
insights into the design and implementation of effective monetary policy. Further studies could
also focus on monetary policy effects on sectoral divisions of the economy as well as other
macroeconomic fundamentals not captured in this study.

104
References
Abiad A., Oomes M. & Ueda K. (2010). The quality effect: Does financial liberalization improve
the allocation of capital? In Crowe C., Johnson S., Ostry D. J. et al., (Eds), International Monetary
Fund, Washington, DC.

Acharya, V. V and Naqvi, H. (2012). The seeds of a crisis: a theory of bank liquidity and risk-
taking over the business cycle. Journal of Financial Economics, 106 (2), 349 – 366.

Adjasi, C. K. D. & Biekpe, N. (2009). Do stock markets matter in investment growth in Africa?
The Journal of Developing Areas, 43 (1), 109-120.

Adjasi, C. K. D. & Biekpe, N. (2006). Stock market development and economic growth: The
case of selected African countries. African Development Review, 18(1), 144–161.

Adolfson, M, 2001. Monetary Policy with Incomplete Exchange Rate Pass-Through. Working
Paper Series, No.127, Sveriges Riksbank.

Adrian, T. & Shin, H. S. (2009). Money, liquidity and monetary policy. American Economic Review,
Papers and Proceedings, 99, 600-05.

Aghion, P., Bacchetta, P., Ranciere R. & Rogoff, R. (2006). Exchange Rate Volatility and
Productivity Growth: The Role of Financial Development NBER Working Paper 12117.
Massachusetts, NBER Inc.

Aghion, P., Harris, C. & Vickers, J. (1997). Competition and growth with step by-step
innovation: An example. European Economic Review, 41, 771-82.

Agoraki, M. K., Delis, M. D. & Pasiouras F. (2011). Regulations, competition and bank risk-
taking in transition countries. Journal of Financial Stability, 7, 38–48.

Agur, I., & Demertzis, M. (2010). Monetary policy and excessive bank risk taking (Discussion
Paper 2010-30S). Tilburg University, Center for Economic Research.

105
Agur, I., & Demertzis, M. (2012). Excessive bank risk-taking and monetary policy. (ECB
Working Paper No. 1457) European Central Bank, Frankfurt.

Ahmed, S. and Islam, Md. E. (2006). Interest rate responsiveness of investment spending in
Bangladesh: A VAR approach. (Working Paper Series, WP 0608, Policy Analysis Unit (PAU)).
Bangladesh: Research Department, Bangladesh Bank.

Akbar, M. & Jamil, F. (2012). Monetary and fiscal policies' effect on agricultural growth: GMM
estimation and simulation analysis. Economic Modelling, 29, 1909–1920.

Allen, F. & Gale, D. (2004). Financial intermediaries and markets, Econometrica, 72, 1023-1061.

Allen, F., & Gale, D. (1999). Bubbles, crises, and policy. Oxford Review of Economic Policy, 15(3), 9-
18.

Altunabas, Y., Gambacorta, L. & Marquez-Ibanez, D. (2010). Does monetary policy affect bank
risk-taking? (BIS Working Paper No. 298). Bank for International Settlement, Basel.

Amarasekara, C. (2008). The impact of monetary policy on economic growth and inflation in Sri
Lanka. CBSL Staff Studies, 38, (1&2).

Angeloni, I., Kashyap, A. K., & Mojon, B. (Eds.). (2003). Monetary policy transmission in the
euro area: a study by the eurosystem monetary transmission network. Cambridge University
Press.

Arellano M., & Bond S. (1991). Some tests of specification for panel data: Monte Carlo evidence
and an application to employment equations. Review of Economic Studies, 58: 287–297.

Asante, Y. (2000). Determinants of Private Investment Behaviour (AERC Research Paper No.
100) Nairobi: AERC.

Aschauer, D.A. (1990). Public Investment and Private Sector Growth. Economic Policy
Institute: Washington, D.C.

106
Athukorala, P. & Sen, K. (2002). Saving, investment and growth in India. New Delhi: Oxford
University Press.

Atje, R. & Jovanic, B. (1993). Stock markets and development. European Economic Review, 37(2–3),
16–41.

Azarmi, T. Lazar, and Jeyapaul, J. 2005. Is the Indian Stock Market a Casino. Journal of Business &
Economics Research, 3(4), 60-72.

Aziz N. (2008). The role of exchange rate in trade balance: Empirics from Bangladesh. Birmingham, UK:
University of Birmingham.

Bagchi, B., Chortareas, G.E. & S.M. Miller (2004). The real exchange rate in small, open
developed economies: Evidence from cointegration analysis, Economic Record, 80, 76-88.

Bagehot, W. (1873). Lombard Street: a description of the money market. Reprinted by Richard D. Irwin
Inc., Homewood, IL, 1962.

Bahmani-Oskooee, M. and S.W. Hegerty (2009) “The Effects of Exchange-Rate Volatility on

Commodity Trade Between the U.S. and Mexico” Southern Economic Journal 79, 1019-1044.

Balassa, B. (1964). The purchasing power parity doctrine: A reappraisal, Journal of Political Economy
72, 584-596.

Baldawi, A. (2005). Private Capital Formation and Macroeconomic Policies in Sudan:

Application of a Simple Cointegrated Vector Autoregressive Model. Unpublished Thesis,
Department of Economics, University of Khartoum.

Baliamoune-Lutz, M., & Ndikumana, L. (2007). The growth effects of openness to trade and the
role of institutions: New evidence from African countries (No. 2007-05). Working Paper,
Department of Economics, University of Massachusetts.

Barbone, L., & Rivera-Batiz F. (1987). Foreign capital and the contractionary impact of currency
devaluation, with an application to Jamaica. Journal of Development Economics, 26, 1-15.

107
Barro R. J. (1991). Economic growth in a cross section of countries. Quarterly Journal of Economics,
106, 407-444.

Barro, R. & Lee, J. (1993). International comparisons of educational attainment. Journal of

Monetary Economics, 32, 363-394.

Barro, R. J. (1990). Government spending in a simple model of endogenous growth. Journal of

Political Economy, 98, 103-125.

Barro, R. J. (1999). Determinants of Democracy. Journal of Political Economy, 107(6), 158-83.

Barro, R. J. (2002). Education as a determinant of economic growth. In E. P. Lazear (ed.)

Education in the twenty-first century (pp. 9 – 24). Stanford: Hoover Institution Press.

Bawumia, M., Owusu-Danso, T., & Arnold, M. (2008). Ghana's reforms transform its financial
sector. IMF Survey Magazine: Countries & Region, 1-4.

Berg, A., & Pattillo, C. (1999). Predicting currency crises: The indicators approach and an
alternative. Journal of International Money and Finance, 18, 561-586.

Berger, A. N., Klapper, L. F. and Turk-Ariss, R. (2009). Bank competition and financial stability.
Journal of Financial Services Research, 35, 99-118.

Berger, A. N. and Bouwman, C. H. S., (2012). Bank Liquidity Creation, Monetary Policy, and
Financial Crises. Available at SSRN: http://ssrn.com/abstract=1952728.

Bergvall, A. (2004). What determines real exchange rates? The Nordic Countries. Scandinavian
Journal of Economics, 106, 315-337.

Bernanke, B., & Gertler, M. (2000). Monetary policy and asset price volatility (No. w7559).
National Bureau of Economic Research Inc., Massachusetts.

108
Bernanke B. S. and Gertler M. (1995). Inside the black Box: the credit channel of monetary
policy transmission, NBER Working Papers 5146, National Bureau of Economic Research, Inc.,
Massachusetts.

Berthelemy, J. C., and Varoudakis, A. (1996). Economic growth, convergence clubs, and the role
of financial development. Oxford Economic Papers, 48(2), 300 – 328.

Bhattacharya, J., Haslag, J., & Martin, A. (2009). Optimal monetary policy and economic growth.
European Economic Review, 53(2), 210-221.

Bhattacharya, P. C., and Sivasubramanian, M. N. (2003). Financial development and economic

growth in India: 1970–1971 to 1998–1999. Applied Financial Economics, 13(12), 925 – 929.

Bhattacharya, R., Patnaik I. and Shah A. (2011). “Monetary Policy Transmission in an Emerging
Market Setting”, IMF Working Paper, 11/5, International Monetary Fund. International Settlements
Quarterly Review, 29-46.

Bhide, A. (1993). The Hidden Costs of Stock Market Liquidity. Journal of Financial Economics, 34,
31–51.

Bikker, J. A., & Metzemakers, P. A. (2005). Bank provisioning behaviour and procyclicality.
Journal of International Financial Markets, Institutions and Money, 15(2), 141-157.

Blomstrom, M., Lipsey, R. E. & Zejan, M. (1996). Is fixed investment the key to economic
growth. The Quarterly Journal of Economics, 111, 269-276.

Blundell, R & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data
models. Journal of Econometrics, 87, 115-143.

Borio, C. E., & Drehmann, M. (2009). Assessing the risk of banking crises–revisited. BIS
Quarterly Review, March.

109
Borio C. & Zhu H. (2008). Capital regulation, risk-taking and monetary policy: A missing link in
the transmission mechanism? (BIS Working Paper, No. 268). Bank for International Settlements,
Basel.

Borio, C, Furfine, C & Lowe, P. (2001). Procyclicality of the financial system and financial
stability: Issues and policy options in Marrying the macro- and micro-prudential dimensions of
financial stability. (BIS Papers, no 1) Bank for International Settlements, Basel.

Borio, C. (2000). Market liquidity and stress: Selected issues and policy implications. BIS
Quarterly Review. (December), 43 – 53.

Boyd, J. H. & De Nicoló, G. (2005). The theory of bank risk taking and competition revisited.
Journal of Finance, 60, 1329-1343.

Boyd, J. H., De Nicoló, G. & Alal, A. M. (2009). Bank competition, risk, and asset allocations
(Working Paper No. 09/143) Washington D.C., International Monetary Fund.

Buffie, E. (1984). Financial Repression, the New Structuralists and Stabilization Policy in Semi-
Industrialized Economies. Journal of Development Economics, 14, 305-322.

Camara, B., Lepetit, L. & Tarazi, A. (2012) Ex ante capital position, changes in the different
components of regulatory capital and bank risk. Available at SSRN:
http://ssrn.com/abstract=1496818 or http://dx.doi.org/10.2139/ssrn.1496818.

Carranza, L., Galdon-Sanchez, J.E. and Biscarri, J.G. (2010) “Understanding the Relationship
between Financial Development and Monetary Policy” Review of International Economics, 18(5),
849–864.

Cecchetti, S. G. (1999). Legal structure, financial structure, and the monetary policy transmission
mechanism (No. w7151). National Bureau of Economic Research, Massachusetts.

Chakraborty, I. (2008). Does financial development cause economic growth? The case of India.
South Asia Economic Journal, 9(1), 109-139.

110
Chami R., Barajas, A., Cosimano, T., Fullenkamp, C., Gapen, M. &` Montiel, P. (2008).
Macroeconomic consequences of remittances (IMF Occasional Paper Series. No. 259). IMF.

Chari, V. V., Jones, L. E., & Manuelli, R. E. (1995). The growth effects of monetary policy.
Federal Reserve Bank of Minneapolis. Quarterly Review-Federal Reserve Bank of Minneapolis,
19(4), 18.

Chatelain J.- B., Ehrmann, M., Generale, A., Martínez-Pagés, J., Vermeulen P., & Worms, A.
(2003a), Monetary policy transmission in the euro area: New evidence from micro data on firms
and banks. Journal of the European Economic Association, 1, 2-3.

Chen C.R., Steiner T. L., & Whyte, A. M. (2006). Does stock option-based executive
compensation induce risk-taking? An analysis of the banking industry. Journal of Banking and
Finance, 30, 915-945.

Chiesa, G. (2001). Incentive-Based Lending Capacity, Competition and Regulation in Banking.

Journal of Financial Intermediation, Elsevier, 10(1), 28 – 53.

Chowdhury, A. R., Fackler, J. S., & McMillin, W. D. (1986). Monetary policy, fiscal policy, and
investment spending: An empirical analysis. Southern Economic Journal, 794-806.

Christiano, L. J., Eichenbaum, M., & Evans, C. L. (2005). Nominal rigidities and the dynamic
effects of a shock to monetary policy. Journal of political Economy, 113(1), 1-45.

Claessens, S. & L. Laeven (2004), What drives bank competition? Some international evidence.
Journal of Money, Credit and Banking, 36, 563 – 583.

Combes, Jean-Louis, Kinda, T. & Plane, P. (2010) Capital Flows, Exchange Rate Flexibility, and
the Real Exchange Rate. (IMF Working Paper, WP 11/9). IMF

Connolly, M. & Da Silviera, J. D. (1979). Exchange market pressure in postwar Brazil: an

application of the Girton-Roper monetary model. American Economic Review, 69, 448 – 54.

111
Cooper, R. N. (1971), Currency Devaluation in Developing Countries. Essays in International
Finance, 86.

Cottarelli, C., & Kourelis, A. (1994). Financial structure, bank lending rates, and the transmission
mechanism of monetary policy. Staff Papers-International Monetary Fund Staff Papers, 41, 587-
623.

Coutinho, R. & Gallo, G. (1991). Do public and private investment stand in each other’s way.
World development report background paper, mimeo. World Bank. Washington, D.C.

Crouhy, M., Galai, D. & Mark, R. (2001). Risk Management. McGraw-Hill, New-York.

De Bondt, G. (2005). Interest rate pass-through: empirical results for the euro area, German
Economic Review, 6, 37–78.

De Bondt, G. J. (2000). Financial Structure and Monetary Transmission in Europe: A Cross Country
Study, UK. Edward Elgar.

De Gregorio, J. & Guidotti, P.E., (1995). Financial development and economic growth. World
Development, 23(3), 433 – 448.

De Long, J. B. & Summers, L. H. (1991). Equipment investment and economic growth. The
Quarterly Journal of Economics, CVI, 445 – 502.

Deelchand, T. and Padgett, C. (2009) The relationship between risk, capital and efficiency:
evidence from Japanese Cooperative Banks (ICMA Centre Discussion Papers in Finance
DP2009-12).University of Reading, ICMA

Delis, M. D., & Kouretas, G. P. (2011). Interest rates and bank risk-taking. Journal of Banking &
Finance, 35(4), 840 – 855.

Dell’Ariccia, G. & Marquez, R. (2006). Lending booms and lending standards. Journal of Finance,
61(5), 2511 – 2546.

112
Dell’Ariccia, G., Laeven, L. & Marquez, R. (2010). Monetary policy, leverage, and bank risk-
taking (IMF Working Paper No. 10/276). IMF.

Demetriades, P. O. & Hussein, K. A. (1996). Does financial development cause economic

growth? Time-series evidence from 16 countries. Journal of development Economics, 51(2), 387-411.

Demirgüç-Kunt, A. & Levine, R. (1996). Stock market development and financial intermediaries:
Stylized facts. The World Bank Economic Review, 10(2), 291–321.

Demirgüç-Kunt, A. and Huizinga, H., (2010). Bank activity and funding strategies: The impact
on risk and returns. Journal of Financial Economics 98 (3), 626 – 650.

Demsetz, R.S., and Strahan, P., (1997). Diversification, size, and risk at bank holding companies.
Journal of Money, Credit and Banking 29, 300 – 313.

Devereux, M, 2001. Monetary Policy, Exchange Rate Flexibility and Exchange Rate Pass
Through. In Revisiting the Case for Flexible Exchange Rates. Bank of Canada.

Devereux, M. B., & Smith, G. W. (1994). International risk sharing and economic growth.
International Economic Review, 35 (4), 535 – 550.

DFID (2004), Financial sector development: A pre‐requisite for growth and poverty reduction?
London Department for International Development (DFID), Policy Division.

Diamond, D. W. (1984). Financial intermediation and delegated monitoring. Review of Economic

Studies, 51, 393–414.

Diamond, D. W. and Rajan, G. R. (2006) Money in a Theory of Banking. American Economic

Review, 96(1), 30-53.

Dickey, D. and Fuller, W. (1979). Distribution of the estimators for autoregressive time series
with a unit root. Journal of the American Statistical Association, 74(366), 427–431.

113
Dieckmann, O. (1996). Cultural Determinants of Economic Growth: Theory and Evidence
Journal of Cultural Economics, 20, 297-320.

Dornbusch, R. (1988) The European monetary system, the Dollar and Yen, in The European
Monetary System, Giavani, F., Micossi, S. and Miller, M. (Eds), Cambridge Press, Cambridge.

Drehmann, M. & Nikolaou, K. (2010). Funding liquidity risk: Definition and measurement (ECB
WP, no. 1024). European Central Bank, Frankfurt.

Easterly, W. & Levine, R. (1997) Africa‘s growth tragedy: Policies and ethnic divisions. Quarterly
Journal of Economics, 112(4) 1203 – 1249.

Easterly, W. & Rebelo, S. (1993). Fiscal policy and economic growth: An empirical investigation.
Journal of Monetary Economics, 417 – 458.

Easterly, W. & Rebelo, S. (1993). Fiscal policy and economic growth: An empirical investigation.
Journal of Monetary Economics, 32(3) 417 – 458.

Egert, Balasz, Imed Drine, Kirsten Lommatzsch, and Christophe Rault. (2006) “The Balassa-
Samuelson Effect in Central and Eastern Europe: Myth or Reality?” Journal of Comparative
Economics, 31, 552–72.

Eichengreen, B., & Arteta, C. (2000). Banking crises in emerging markets: Presumptions and
evidence. (Center for International and Development Economics Research Paper C00-115).
Berkeley, CA: University of California at Berkeley.

Eichengreen, B., Rose, A. J. C, & Wyplosz, C (1995). Exchange market mayhem: the antecedents
and aftermath of speculative attacks. Economic Policy, 21, 249-31.

Emery, R. (1971): The use of interest rate policies as a stimulus to econometric growth
(International Finance Discussion Papers #676 in Review of Foreign Developments (RFD)
series, Federal Reserve Board, Washington D.C.

114
Engle, R. and Granger, C. (1987), Co-integration and error correction: representation, estimation
and testing, Econometrica, 55(2), 251–276.

Erceg, C., & Levin, A. (2006). Optimal monetary policy with durable consumption goods. Journal
of Monetary Economics, 53(7), 1341 – 1359.

Fan, F., Yu, Y. & Zhang C. (2011). An empirical evaluation of China’s monetary policies Journal
of Macroeconomics, 33, 358 – 371.

Fedderke, J. W. (2005). Technology, human capital and growth (No. 27). Economic Research
Southern Africa.

Feridun, M. (2009). Determinants of exchange market pressure in Turkey: An econometric

investigation. Emerging Markets Finance and Trade, 45, 65-81.

Filardo, A. J. (2000). Monetary policy and asset prices. Economic Review-Federal Reserve Bank of
Kansas City, 85(3), 11-38.

Filer, R. K., Hanousek, J., & Campos, N. F. (1999). Do stock markets promote economic
growth? (Working Paper No. 26). Centre for Economic Research and Graduate Education.
Prague: Charles University.

Frenkel, R. and L. Taylor (2006). “Real Exchange Rate, Monetary Policy and Employment,”
DESA Working Paper 19. United Nations. New York.

Fry M. J. (1997a). Emancipating the banking System and developing markets for government
debt. London: Routledge.

Fry, M. J. (1980). Money, interest, inflation and growth in Turkey. Journal of Monetary Economics, 6,
535-545.

Fry, M. J. (1988). Money, interest, and banking in economic development. Baltimore: Johns Hopkins
Press.

115
Furlong, F.T. & Keely, M.C. (1990). Capital Regulation and Bank Risk Taking: A note. Journal of
Banking and Finance, 13, 883-891.

Gaiotti, E., & Generale, A. (2001). Does monetary policy have asymmetric effects? A look at the
investment decisions of Italian firms. European Central Bank, Working Paper Series, No. 110.

Gamber, E. N., & Hakes, D. R. (2005). Is monetary policy important for forecasting real growth
and inflation? Journal of Policy Modeling, 27(2), 177-187.

Galbis, V. (1979). Money, investment and growth in Latin America, 1961-1973. Economic
Development and Cultural Change, 27: 423 – 43.

Gali, J., & Monacelli, T. (2005). Monetary policy and exchange rate volatility in a small open
economy. The Review of Economic Studies, 72(3), 707 – 734.

Gambacorta, L and P E Mistrulli (2004): “Does bank capital affect lending behaviour?” Journal of
Financial Intermediation, Vol. 13, 436 – 57.

Gambacorta, L. (2009). Monetary Policy and the Risk Taking Channel. BIS Quarterly Review,
December, Bank for International Settlements.

Garcia Herrero, A., & López, P. D. R. (2003). Implications of the design of monetary policy for financial
stability. EconWPA.

Garcia, C. & N. Malet (2007). Exchange market pressure, monetary policy, and economic
growth: Argentina, 1993-2004. Developing Economies, 45, 253 – 282.

Gelb, A. H. (1989). Financial Policies, Growth and Efficiency (Working Papers WPS 202).
Washington, D.C: The World Bank.

Geromichalos, A., Licari, J. M., & Suárez-Lledó, J. (2007). Monetary policy and asset prices.
Review of Economic Dynamics, 10(4), 761 – 779.

116
Gigineishvili, N. (2011). Determinants of the interest rate pass-through: Do macroeconomic
conditions and financial market structure matter? (IMF Working Paper 11/176). IMF

Gilchrist, S., & Leahy, J. V. (2002). Monetary policy and asset prices. Journal of monetary Economics,
49(1), 75 – 97.

Girton, L. & Roper, D. (1977). A monetary model of exchange market pressure applied to the
post-war Canadian experience. American Economic Review, 67, 537 – 548.

Gochoco-Bautista, M.S. & Bautista, C. C. (2005). Monetary policy and exchange market
pressure: The case of the Philippines. Journal of Macroeconomics, 27, 153-168.

Goetz, M. R., Laeven, L. & Levine, R. (2011). The valuation effects of geographically diversified
bank holding companies: Evidence from U.S. Banks (NBER Working Paper No. 17660).
Massachusetts, National Bureau of Economic Research Inc.

Goldfajn, I. & Baig, T. (1998). Monetary policy in the aftermath of currency crisis: The case of
Asia (IMF Working Paper 98/170). Washington D.C: IMF.

Gonzales, R., (2005), Bank regulation and risk-taking incentives: An international comparison of
bank risk. Journal of Banking and Finance, 29, 1153-1884.

Guitian, M. (1976). The effects of changes in the exchange rate on output, prices, and the
balance of payments. Journal of International Economics, 6, 65 – 74.

Gunsel N. R., Tursoy, T. & Rjoub, H. (2010). An empirical analysis of currency crises,
fundamentals and speculative pressure. African Journal of Business Management, 4, 972- 978.

Gupta, R., & Kabundi, A. (2010). The effect of monetary policy on house price inflation: a factor
augmented vector autoregression (FAVAR) approach. Journal of Economic Studies, 37 (6), 616 –
626.

117
Gutiérrez, M. A. (2005). Economic growth in Latin America: the role of investment and other
growth sources. Economic Development Division, CEPAL-SERIE Macroeconomía del
desarrollo No. 36.

Hansen, B. E. & Seshadri, A. (2013). Uncovering the relationship between real Interest rates and
economic growth (Research Policy Brief No. 303). Michigan: University of Michigan Retirement
Research Centre.

Harrison, L. E. (1992). Who Prospers? How Cultural Values Shape Economic and Political Success. New
York, NY: Harper Collins, Basic Books.

Hegerty, S. W. (2010). Exchange-market pressure and currency crises in Latin America:

Empirical tests of their macroeconomic determinants. Economics Bulletin, 30, 2210-2219.

Hellman, T., Murdock, K. & Stiglitz, J. E. (2000). Liberalization, moral hazard in banking and
prudential regulation: Are capital controls enough? American Economic Review, 90, 147- 165.

Hendry, D.F. (2005), Dynamic Econometrics, Oxford University Press, Oxford.

Hernandez-Cata, E. (2000). Raising growth and investment in Sub-Saharan Africa: What can be
done? (Policy Discussion Paper: PDP/00/4) Washington, D .C: International Monetary Fund.

Hicks, J. R. (1969). Critical Essays in Monetary Theory. Oxford: Oxford University Press.

Hoti, A. H., & Nuhiu, A. R. (2011). Effects of capital markets development on economic growth
of Western Balkan countries. European Journal of Economics, Finance and Administrative Sciences, (43),
88.

Hsiao, C. (2003). Analysis of Panel Data (2nd ed). Cambridge: Cambridge University Press.

Huang, Z., (2003). Evidence of bank lending channel in the UK. Journal of Banking and Finance, 27,
491 – 510.

118
Hughes, J. P., Mester, L. J., & Moon, C. G. (2001). Are scale economies in banking elusive or
illusive?: Evidence obtained by incorporating capital structure and risk-taking into models of
bank production. Journal of Banking & Finance, 25(12), 2169-2208.

Hülsewig, H., Mayer, E. & Wollmershäuser, T. (2006). Bank loan supply and monetary policy
transmission in Germany: An assessment based on matching impulse responses. Journal of
Banking and Finance, 30, 2893 – 2910.

Iacoviello, M. & Minetti, R., (2008). The credit channel of monetary policy: evidence from the
housing market. Journal of Macroeconomics, 30, 69–96.

Iacoviello, M. (2000) House prices and the macro economy in Europe: Results from a structural
VAR analysis. ECB Working Paper, No. 18.

Iacoviello, M., & Minetti, R. (2003). Financial liberalization and the sensitivity of house prices to
monetary policy: Theory and evidence. The Manchester School, 71(1), 20-34.

Iacoviello, M. (2005). House prices, borrowing constraints, and monetary policy in the business
cycle. American economic review, 739-764.

IMF. (2009). De Facto Classification of Exchange Rate Regimes and Monetary Policy
Frameworks. IMF. Retrieved from:
https://www.imf.org/external/np/mfd/er/2008/eng/0408.htm.

Ioannidou, V., Ongena, S., & Peydró, J. L. (2015). Monetary policy, risk-taking, and pricing:
Evidence from a quasi-natural experiment. Review of Finance, 19(1), 95-144.

Ioannidou, V., Ongena, S., & Peydró, J. L. (2009). Monetary policy and subprime lending: a tall
tale of low federal funds rates, hazardous loans and reduced loan spreads. European banking
centre discussion paper, 45.

Iwatsubo, K. (2007). Bank capital shocks and portfolio risk: Evidence from Japan. Japan and the
World Economy, Elsevier, 19(2), 166 – 186.

119
Izquierdo, A. & Montiel, P. J. (2006). Remittances and equilibrium real exchange rates in six
Central American Countries. (unpublished thesis), Williamstown, Massachusetts: Williams
College).

Jayaraman, T. K. & Chee-Keong, C. (2008). Exchange market pressure in a small Pacific Island
country: A study of Fiji: 1975-2005. International Journal of Social Economics, 35, 985 – 1004.

Jimenez, G., Ongena, S., Peydro, J. -L. & Saurina, J. (2008). Hazardous times for monetary
policy: What do twenty-three million bank loans say about the effects of monetary policy on
credit risk-taking? (Bank of Spain WP 0833). Spain: Bank of Spain.

Johansen, S., (1988), “Statistical Analysis of Cointegration Vectors,” Journal of Economic Dynamics
and Control, Vol. 12, No. 2–3, 231–254.

John, M., & Pokhariyal, G. P. (2013). Two stage interest rate pass through in Kenya. Journal of
Emerging Trends in Economics and Management Sciences, 4(1), 54.

Johnson, H. (1967). Essays in monetary economics. Cambridge, Harvard University Press.

Kahn, B. (1998). Inflation Targeting. CREFSA, London School of Economics, Quarterly Review,
2, 17-25.

Kamaly, A. & Erbil, N. (2000). A VAR analysis of exchange market pressure: A case study for
the MENA region (ERF Working Paper no. 200025). Cairo: Economic Research Forum.

Kaminsky, G., Lizondo, S., & Reinhart, C. M. (1998). Leading indicators of currency crises. Staff
Papers-International Monetary Fund, 45 (1) 1-48.

Kar, M., Peker, O., & Kaplan, M. (2008). Trade liberalization, financial development and
economic growth in the long term: The case of Turkey. South East European Journal of Economics
and Business, 3(2), 25-38.

Karfakis, C. & Moschos, D. (2004). Predicting currency crises: Evidence from two transition
economies. Emerging Markets Finance and Trade, 40, 95-103.

120
Kashyap, A. K. and Stein, J. C. (1995). The Impact of Monetary Policy on Bank Balance Sheets.
Carnegie-Rochester Conference Series on Public Policy, 42: 151-195.

Kashyap, A. K., Stein, J. C. & Wilcox, D. W. (1993). Monetary policy and credit conditions:
Evidence from the composition of external finance. American Economic Review, 83(1): 78 – 98.

Katircioglu, S.T., Kahyalar, N. & Benar, B. (2008). Financial development, trade and growth
triangle: the case of India. International Journal of Social Economics, 37 (9), 586-98.

Keely, M. (1990). Deposit insurance, risk, and market power in banking. American Economic
Review, 80, 1184–1200.

Kehl J. R. (2007) Emerging markets in Africa. African Journal of Political Science and International
Relations, 1(1), 001-008.

Kelly, T. (1997). Public expenditures and growth Journal-of-Development-Studies, 4, 60-68.

Khan, A. (2001). Financial development and economic growth. Macroeconomic Dynamics, 5(03),
413-433.

Khan, I. N. (2010). Exchange market pressure index in Pakistan (SBP Working Paper Series No.
35). State Bank of Pakistan.

Khan, M. S. and Senhadji, A. S. (2000), "Financial Development and Economic Growth: An

Overview", IMF Working Paper, WP/00/209. 27.

Khumalo, J. (2014.) Interest Rate – Private Capital Formation Nexus in South Africa: Bounds
Test Approach. Mediterranean Journal of Social Sciences, 5(1), 25 – 38.

Kibritcioglu, B., Kose, B. & Ugur, G. (1998). A leading indicators approach to the predictability
of currency crises: The case of Turkey. Hazine Dergisi, 12, 1–27.

121
Kiendrebeogo, Y. (2012). The effects of financial development on trade performance and the
role of institutions, Economics Bulletin, 32(3), 2546-2559.

Kim, D. & Santomero, A. (1988). Risk in banking and capital regulation. Journal of Finance 43,
1219–1233.
King, R. G., & Levine, R. (1993). Finance, entrepreneurship and growth. Journal of monetary
Economics, 32(3), 513-542.

Koehn, M. & Santomero, A. (1980). Regulation of bank capital and portfolio risk. Journal of
Finance, 35, 1235–1244.

Kovanen, A. (2011) Monetary Policy Transmission in Ghana: Does the Interest Rate Channel
Work? IMF Working Papers WP/11/275, International Monetary Fund.

Krause, S., & Rioja, F. (2006). Financial development and monetary policy efficiency. Emory
Economics, 613.

Krugman, P. & Taylor, L., (1978). Contractionary effects of devaluation. Journal of International
Economics, 8, 445–456.

Kumar, M. & Mlambo, K. (1995). Determinants of Private Investment in Sub-Saharan Africa:

An Empirical Analysis. A paper Presented at the 11th World Congress of the International
Economic Association, Tunis, December: 22-24.

Kwan, S. & Eisenbeis R. (1997). Bank risk, capitalization, and operating efficiency. Journal of
Financial Services Research, 12(2), 117-131.

Laeven, L. & Levine, R., (2009). Bank governance, regulation and risk taking. Journal of Financial
Economics, 93, 259–275.

Laeven, L., & Majnoni, G. (2003). Loan loss provisioning and economic slowdowns: too much,
too late? Journal of financial intermediation, 12(2), 178-197.

122
Lartey E., (2007). Capital inflows and the real exchange rate: An empirical study of Sub-Saharan
Africa. Journal of International Trade & Economic Development, 16, 337 – 357.

Lastrapes, W. D., & Douglas McMillin, W. (2004). Cross‐Country variation in the liquidity effect:
The role of financial markets. The Economic Journal, 114 (498), 890 – 915.

Levine, R. (1997), “Financial Development and Economic Growth: Views and Agenda”, Journal
of Economic Literature, Vol. 35, 688 – 726.

Levine, R., & Zervos, S. (1996). Stock market development and long-run growth. World Economic
Review, 10 (2), 323 – 339.

Levine, R., Loayza, N. and Beck, T., (2000). Finance and the Sources of Growth. Journal of
financial economics, 58 (1), 261 – 300.

Licchetta, M. (2009) Common determinants of currency crises: role of external balance sheet
variables. Bank of England, Working Paper No. 366.

Lipsey R. G. & Chrystal, K. A. (2011), 'Economics', (12th Ed.) Oxford University Press, Oxford.

Lown, C. & Morgan, D. P. (2006). The credit cycle and the business cycle: New findings using
the Loan Officer Opinion Survey. Journal of Money, Credit and Banking, 38(6), 1575 – 1597.

Lucas, R. E. (1988). On the Mechanics of Economic Development. Journal of Monetary Economics,

22 (1), 3 – 42.

Lueth, E., & Ruiz-Arranz, M. (2007). Are workers‘ remittances a hedge against macroeconomic
shocks? The case of Sri Lanka (IMF Working Paper 07/22) Washington: International Monetary
Fund.

M’Amanja, D. & Morrissey, O. (2006). Foreign Aid, Investment, and Economic Growth in
Kenya: A Time Series Approach. Available online at:
https://www.nottingham.ac.uk/credit/documents/papers/06-05.pdf

123
Maddaloni, A. & Peydro, J.L. (2010). Bank risk taking, securitization, supervision, and low
interest rates: Evidence from lending standards. Review of Financial Studies, 24 (6), 2121–2165.

Mandilaras A. & Bird G. (2008). Foreign exchange pressures in Latin America: Does debt
Matter? Journal of International Development, 20, 613–627.

Mangani, R. (2012). “The Effects of Monetary Policy on Prices in Malawi.” AERC Research
Paper 252.

Masten, A. B., Coricelli, F., and Masten, I. (2008). Non-linear growth effects of financial
development: Does financial integration matter? Journal of International Money and Finance, 27(2),
295 – 313.

Matsuyama, K. (2007). Credit Traps and Credit Cycles. American Economic Review, 97 (1), 503 –
516.

Matwang, F. (2000). “Determinants and constraints to private investment: The case of Kenya”,
African Institute for Economic Development and Planning (IDEP).

Mazumdar, J. (1996). Do static gains from trade lead to medium-run growth?. Journal of Political
Economy, 104(6), 1328-1337.

McCarthy, J., & Peach, R. W. (2002). Monetary policy transmission to residential investment.
Federal Reserve Bank of New York Economic Policy Review, 8 (1), 139 – 158.

McCleary, R. M. & Barro, R. J. (2006). Religion and Economy. Journal of Economic Perspectives, 20
(2), 49 – 72.

McKinnon, R. I. (1973). Money and capital in economic development. Washington: Brookings

Institution.

McPherson, M. F. & Rakovski, T. (2000). Exchange rates and economic growth in Kenya: An
econometric analysis African Economic Policy (Discussion Paper No. 56). EAGER Publishers,
Alexandria, VA.

124
Mehran, H. & Rosenberg, J. (2008). The effects of employee stock options on bank investment
choice, borrowing, and capital (Reports no. 305). Federal Reserve Bank of New York.

Mehrara, M. & Karsalari, A. R. (2011). The nonlinear relationship between private investment
and real interest rates based on dynamic threshold panel: The case of developing countries.
Journal of Money, Investment and Banking, 21, 32 – 42.

Meier, A. and Muller, G. J (2005) Fleshing out the Monetary Transmission Mechanism: Output
Composition and the Role of Financial Frictions, Working Paper Series 0500, European Central
Bank.

Mercieca, S., Schaeck, K., & Wolfe, S. (2007). Small European banks: Benefits from
diversification?. Journal of Banking & Finance, 31(7), 1975-1998.

Miller, M. (1988). Financial Markets and Economic Growth. Journal of Applied Corporate Finance,
Morgan Stanley, 11 (3), 8 – 15.

Minella, Andre, (2001), “Monetary Policy and Inflation in Brazil (1975−2000): A VAR
Estimation,” Banco Central do Brasil Working Paper No. 33 (November).

Mishkin, F. (1996): The Channels of Monetary Transmission: Lessons for Monetary Policy.
(NBER Working Paper 5464). National Bureau of Economic Research, Inc. Massachusetts.

Mishkin, F. (2007). Will Monetary Policy Become More of a Science? (NBER Working Papers
13566). National Bureau of Economic Research, Inc. Massachusetts.

Mishra, P. & Montiel, P. (2013). How effective is monetary transmission in low-income

countries? A survey of the empirical evidence. Economic Systems, 37 (2), 187 – 216.

Mishra, P., Montiel, P., & Spilimbergo, A. (2011). How effective is monetary transmission in
developing countries? A survey of the empirical evidence. London, Centre for Economic Policy
Research.

125
Mojon, B. (2000). Financial structure and the interest rate channel of ECB monetary policy
(Working Paper No. 40). Frankfurt, European Central Bank.

Monacelli, T, 2005. Monetary Policy in a Low Pass-Through Environment. Journal of Money Credit
and banking (37), 1047 – 1066.

Moreno, R, (2008). Monetary Policy Transmission and Long Term Interest Rate in Emerging
Markets. In Transmission Mechanism for Monetary Policy in Emerging Market Economies. No.
35, BIS.

Morgan, D. (1993) Asymmetric effects of monetary policy. Federal Reserve Bank of Kansas City
Economic Review, 78, 21–33.

Mukherjee, S. and Bhattacharya, R. (2011). “Inflation targeting and monetary policy transmission
mechanisms in emerging market economies”, IMF Working Paper No. WP/11/229,
International Monetary Fund.

Munir, R., Awan U.R. & Hussain, Z. (2010) Investment, Savings, Interest Rate and Bank Credit
to the Private Sector Nexus in Pakistan. International Journal of Marketing Studies, 2(1) 140 – 146.

Munnell, A. (1992). Infrastructure investment and economic growth. Journal of Economic

Perspectives, 6, 189 – 198.

Mwega, F. M. (2003). Financial Sector Reform in Eastern and Southern Africa. In T.

Mkandawire & C. Soludo (Eds), African Voices on Structural Adjustment in Sub-Saharan Africa: Third
World Press.

Myers, S. C. & Rajan, R. G. (1998). The paradox of liquidity. Quarterly Journal of Economics, 113 (3),
733 – 771.

Nouri, M. & Samimi, A. J. (2011). The Impact of Monetary Policy on Economic Growth in
Iran” Middle-East Journal of Scientific Research, 9(6), 740 – 743.

126
Odedokun, M. O. (1996). Alternative econometric approaches for analyzing the role of the
financial sector in economic growth: Time-series evidence from LDCs. Journal of Development
Economics, 50(1), 119 – 146.

Ogunbenro, I.O., Alawode, A.A. & Mutabuka, P.C. (1996). The effects of a liberalized financial
system on savings, investment, and growth in Nigeria: 1983 – 1996. African Journal of Economic
Policy, June, 27 – 40.

Opolot, J., Nampewo, D., Ntumwa, C and Nyanzi, S. (2013) Financial Architecture and the
Monetary Policy Transmission Mechanism in Uganda Working Paper Series No. 02/2013, Bank
of Uganda.

Parlaktuna, I. (2005). Exchange market pressure in Turkey, 1993–2004: An application of the

Girton-Roper Monetary Model. International Economic Journal, 19, 51–62.

Pesaran, M.H., Shin, Y.C. & Smith, R. (2001). Bound testing approaches to the analysis of level
relationships. Journal of Applied Econometrics, 16, 289–326.

Pétursson, G. (2001). The transmission mechanism of monetary policy: Analysing the financial
market pass-through. Central Bank of Iceland Working Papers, No. 14.

Phillips, P. & Perron, P. (1988). Testing for a unit root in time series regression. Biometrika, 75
(2), 335–346.

Rabanal, P. (2007). Does inflation increase after a monetary policy tightening? Answers based on
an estimated DSGE model. Journal of Economic Dynamics and Control, 31(3), 906 – 937.

Radelet, S. (2010). Emerging Africa: How 17 countries are leading the way. Washington: Center
for Global Development.

Rajan R.G. (2005). Has financial development made the world riskier? (NBER Working Paper
Series, No. 11728). National Bureau of Economic Research Inc. Massachusetts

127
Rajan, R. G., & Subramanian, A. (2005). What undermines aid's impact on growth? (NBER
Working Paper Series, No. w11657). National Bureau of Economic Research Inc. Massachusetts.

Raveena, F. and Natalucci, F. (2008) Monetary Policy Choices in Emerging Market Economies:
The Case of High Productivity Growth. Journal of Money, Credit and Banking, Vol. 40, No. 2–3, 243
– 271.
Rebelo, S. (1991). Long run policy analysis and long-run growth. Journal of Political Economy, 99
(3), 500 – 521.

Reed, R. R. & Ghossoub, E. A. (2012). The effects of monetary policy at different stages of
economic development. Economic Letters, 117, 138 – 141.

Reinhart, C. M., & Rogoff, K. S. (2008). Is the 2007 US sub-prime financial crisis so different?
An international historical comparison. American Economic Review, 98 (2), 339 – 44.

Repullo, R. (2004). Capital requirements, market power, and risk-taking in banking. Journal of
Financial Intermediation, 13, 156 – 182.

Ribeiro, M. B. and Teixeira, J. R. (2001). An econometric analysis of private investment in Brazil.

CEPAL Review. 74, 153 – 166.

Rigobon, R., & Sack, B. (2004). The impact of monetary policy on asset prices. Journal of Monetary
Economics, 51 (8), 1553 – 1575.

Robinson, J. (1952). The Generalization of the General Theory: The Rate of Interest and other Essays.
London. UK: Macmillan Press.

Romer, P.M. (1990). Endogenous technological change. Journal of Political Economy, 98, 71 – 102.

Rosenberg, M.R. (1996) Currency Forecasting; A Guide to Fundamental and Technical Models
of Exchange Rate Determination. Irwin Professional Publishing.

Roubini, N. & Sala-i Martin, X. (1991). Financial Development, Trade Regimes and Economic
Growth (NBER Working Paper no: 3876), NBER Inc. Massachusetts.

128
Rousseau, P.L. & Wachtel, P. (2000). Equity market and growth, cross country evidence on
timing and outcomes 1980-95. Journal of Banking & Finance, 24 (12), 1933 – 57.

Saborowski, C. (2009). Capital inflows and the real exchange rate: Can financial development
cure the Dutch disease? (Working Paper 09/20). IMF.

Salas, V. & Saurina, J. (2003), Deregulation, market power and risk behaviour in Spanish banks.
European Economic Review, 47, 1061 – 1075.

Samba, M. C., and Yan, Y., (2010). “Interest rate pass-through in the Central African economic
and monetary community area: Evidence from an ADRL analysis”. International Journal of Business
Management. 5 (1): 31−41.

Samuelson, P. (1964). Theoretical notes on trade problems. Review of Economics and Statistics, 46,
145 – 154.

Sanchez-Robles, B. (1998). Infrastructure, investment and growth: Some empirical evidence.

Contemporary Economic Policy, 16, 98 – 108.

Sander, H., & Kleimeier, S. (2004). Convergence in euro-zone retail banking? What interest rate
pass-through tells us about monetary policy transmission, competition and integration. Journal of
International Money and Finance, 23 (3), 461– 492.

Senbet, L., & Otchere, I. (2008). African stock markets. In African finance for the 21st Century.
High–level seminar organized by the IMF Institute in collaboration with the Joint Africa
Institute (Tunis–Tunisia; March).

Scharfstein, D. S. and Stein, J. C. (2000). The dark side of internal capital markets: Divisional
rent-seeking and inefficient investment. Journal of Finance, 55, 2537 – 2564.

Schumpeter, J. A. (1912). The Theory of Economic Development, Cambridge. Mass. Havard University
Press.

129
Schumpeter, J. A. (1934). The theory of economic development (R. Opie, Trans.). Cambridge, MA:
Harvard University Press. (Original work published in 1912).

Serven, L. (2002). Real exchange rate uncertainty and private Investment in developing countries
(Domestic Finance Working Paper 2823). Washington, D.C: World Bank.

Serven, L., and A. Solimano, (1990). Private investment and macroeconomic adjustment: Theory,
country experience and policy implications. Washington, D.C: World Bank.

Shafik, N. & Jalali, J. (1991). Are high real interest rates bad for world economic growth? (Policy,
Research, and External Affairs working papers; no. WPS 669). International economic analysis
and prospects. Washington, DC: World Bank.

Shafik, N. (1992). Modelling private investment in Egypt. Journal of Development Economics. 39, 263
– 277.

Shapiro, M. D. (1986). Investment, output, and the cost of capital. Brookings Papers on
Economic Activity, 1, 111 – 152.

Shaw, E. S. (1973). Financial Deepening in Economic Development. New York, NY: Oxford University
Press.

Shleifer, A., & Vishny, R. W. (1986). Large shareholders and corporate control. Journal of Political
Economy, 94 (3), 461 – 88.

Shrieves, R. E. & Drew, D. (1992). The relationship between risk and capital in commercial
banks. Journal of Banking and Finance, 16, 439 – 457.

Singh, A. (1997). Financial liberalization, stock market and economic growth. Economic Journal,
107 (442), 771 – 782.

Singh, S., Razi,A. Endut, N. and Ramlee, H. (2008) Impact of Financial market developments on
the monetary transmission mechanism, Bank of International Settlement, 39, 49 – 99.

130
Smets, F. and Wouters, R, (2002). Openness, imperfect exchange rate pass-through and
monetary policy. Journal of Monetary Economics 49, 947 – 981.

Sørensen, C. K., and Werner, T. (2006). Bank interest rate pass-through in the euro area: a cross country
comparison. Frankfurt am Main: European Central Bank.

Starr, M., (2005). Does money matter in the CIS? Effects of monetary policy on output and
prices. Journal of Comparative Economics, 33, 441 – 461.

Stein, J. C. (1998). An Adverse Selection Model of Bank Asset and Liability Management with
Implications for the Transmission of Monetary Policy, Rand. Journal of Economics, 29, 466 – 486.

Stiglitz, J. E. (1994). The role of the state in financial markets. In: Bruno, M., Pleskovic, B.
(Eds.), Proceedings of the World Bank Annual Conference

Stiglitz, Joseph E., and Andrew Weiss (1981). ‘Credit Rationing in Markets with Imperfect
Information’, American Economic Review, 393 – 410.

Stucka, T. (2004). The effects of exchange rate change on the trade balance in Croatia (Working
Paper No. 65). IMF.

Susanto, D., Rossini, C. P., & Costa, R. (2011). Financial Development and International Trade:
Regional and Sectoral Analysis in 2011 Annual Meeting, July 24-26, 2011, Pittsburgh,
Pennsylvania (No. 102647). Agricultural and Applied Economics Association.

Tanner, E. (2001). Exchange market pressure and monetary policy: Asia and Latin America in
the 1990s. IMF Staff Papers, 47, 311 – 33.

Taylor, J. B. (1994). The inflation-output variability trade-off revisited. In goals, guidelines and
constraints facing monetary policymakers (Series No. 38: 21−38). Boston: Federal Reserve Bank
of Boston Conference.

Taylor, J. B., (2009). The financial crisis and the policy responses: an empirical analysis of what
went wrong (NBER Working Paper No. 14631).Massachusetts, NBER Inc.

131
Tobin, J. (1965). Money and economic growth. Econometrica, 33, 671-84.

Tobin, J. (1969). A general equilibrium approach to monetary theory. Journal of Money Credit and
Banking, 1, 15 – 29.

Tornell, A. and Westermann F. (2002), Boom-Bust cycles in middle income countries: Facts and
explanation (NBER Working Paper Series, No. 9219). National Bureau of Economic Research
Inc., Massachusetts.

Ucer, M., Rijckeghem, C.V. & Yolalan. R. (1998). Leading indicators of currency crises: A brief
literature survey and an application to Turkey. Yapi Kredi Economic Review, 9, 3 – 23.

Uhlig, H. (2005). What are the effects of monetary policy on output? Results from an agnostic
identification procedure. Journal of Monetary Economics, 52 (2), 381 – 419.

Valencia, F., (2011). Monetary policy, bank leverage, and financial stability (IMF Working Paper
11/244). Washington D.C., IMF.

Van den Heuvel, S. J. (2002). Does bank capital matter for monetary transmission? Economic
Policy Review, 8 (1) 259 – 265 Federal Reserve Bank of New York.

Van Els, P., A. Locarno, Morgan, J. & Villetelle, J. P. (2001). Monetary policy transmission in the
euro area: What do aggregate and national structural models tell us? (Working Paper Series No.
94). Frankfurt, European Central Bank.

Van Poeck, A., Vanneste, J. & Veiner, M. (2007). Exchange rate regimes and exchange market
pressure in the new EU member states. Journal of Common Market Studies. 45, 459 – 485.

Van Wijnbergen, S. (1983). Interest Rate Management in LDCs. Journal of Monetary Economics 12 ,
433 – 452.

World Bank (1989). World Development Report: Financial Systems and Development. New York, NY:
Oxford University Press.

132
Wu, D., Yang, J. & Hong, H. (2011). Securitization and Banks’ Equity Risk. Journal of Financial
Services Research. 39, 95 – 117.

Yartey, A. C. (2008). The determinants of stock market development in emerging economies: Is

South Africa different? (IMF Working Paper No. WP/08/32). Washington D.C.: International
Monetary Fund.

Young, A. (1994). Lessons from the east Asian NICS: A contrarian view. European Economic
Review, 38, 3 – 4.

Zakir, N. & Malik, W. S (2013). Are the effects of monetary policy on output asymmetric in
Pakistan? Economic Modelling, 32, 1 – 9.

Zorzi, MC, Hahn, E, Sanchez, M, (2007). Exchange Rate Pass Through in Emerging Markets.
Working Paper Series, No. 739, European Central Bank.

133
 Appendix A
A_1: Variable Definitions
Variable Measurement
LENDING RATE (LR) Average annual lending rate for banking sector
CENTRAL BANK POLICY Annual average of Central Bank policy rate
RATE (MPR)
FINANCIAL Domestic credit to the private sector (% of GDP)
DEVELOPMENT (BSIZE)
ECONOMIC GROWTH Real GDP growth
(GDP)
MARKET VOLATILITY Standard deviation of prime rate (5-year rolling window)
(BVOL)

A_2: Summary Statistics of Study Variables by Country

Gambia
Variable Mean Std. Dev. Min Max
Lr 24.18919 5.636379 14 37
Mpr 14.96432 6.333497 6 29.25
Bsize 12.97703 6.568595 3.73 25.12
GDP 3.752162 3.528698 -4.3 12.39
Bvol 2.628226 2.303575 0 8.871302

Ghana
Variable Mean Std. Dev. Min Max
Lr 27.78378 8.310419 19 47
Mpr 21 10.35105 8 45
Bsize 7.651892 4.944318 1.54 15.88
GDP 3.552703 4.837447 -12.43 14.39
Bvol 3.75164 1.830891 .9797959 8.064738

Nigeria
Variable Mean Std. Dev. Min Max
Lr 16.16177 6.389534 6 31.65

134
Mpr 11.71622 5.246084 3.5 26
Bsize 14.72873 6.798227 6.814002 38.59011
GDP 3.462274 5.227168 -13.12788 10.6
Bvol 2.057864 1.276957 0 5

135
 Appendix B: Regressions on sensitivity to model specifications
B-1: Regression output on complete sample with one (1) lag of GDP
Variable Regression #1 Regression #2

EMP(-1) -0.0264559 (0.000)*** -0.0288086 (0.000)***

MPR -3.950874 (0.081)* -4.658843 (0.018)**

PKFL -8.800449 (0.030)** -6.680622 (0.088)*

DEBT -1.720672 (0.000)*** -1.956307 (0.000)***

TOT - -219.0571 (0.000)***

CABA -4.702457 (0.005)** -

GDP 2.028739 (0.468) 4.532991 (0.443)**

GDP(-1) 0.0984553 (0.973) 2.826173 (0.461)

Wald Chi2, 1812.42 (0.000) 6735.03 (0.000)

Arellano-Bond Test for Order 1 -1.3172 (0.1878) -1.3071 (0.1912)

Autocorrelation
Order 2 0.00046 (0.9996) -0.23292 (0.8158)

Sargan’s Test 11.7107 (1.0000) 12.02655 (1.0000)

***, ** and * denotes significance at 1%, 5% and 10% respectively. Values in parenthesis are the p-values. The
following diagnostic tests are run for the model: the Sargan test for over identification, the Arrelano-Bond test for
autocorrelation and the Wald Chi2 for model fit. The results indicate that the over identifying restrictions are valid,
there is no first order or second order autocorrelation and the model fits the data well.

136
B-2: Regression output on complete sample with two (2) lags of GDP
Variable Regression #1 Regression #2

EMP(-1) -.031608 (0.000)*** -.03199 (0.000)***

MPR --3.863759 (0.184) -2.102238 (0.496)

PKFL 1.810495 (0.804) 8.365 (0.334)

DEBT -1.301327 (0.031)** -1.385754 (0.084)*

TOT - -149.0739 (0.183)

CABA -2.80531 (0.142) -

GDP 8.892914 (0.001)*** 9.288287 (0.015)**

GDP(-1) 6.753634 (0.126) 6.155255 (0.200)

GDP(-2) 10.35197 (0.000)*** 12.80457 (0.000)***

Wald Chi2, 544.00 (0.000) 11196.24 (0.000)

Prob.

Order -1.3112 (0.1898) -1.3079 (0.1909)

Arellano-Bond Test 1
for Autocorrelation Order -0.49858 (0.6181) -0.58664 (0.5574)
2
Sargan’s Test 11.30779 (1.0000) 9.408012 (1.0000)

***, ** and * denotes significance at 1%, 5% and 10% respectively. Values in parenthesis are the p-values. The
following diagnostic tests are run for the model: the Sargan test for over identification, the Arrelano-Bond test for
autocorrelation and the Wald Chi2 for model fit. The results indicate that the over identifying restrictions are valid,
there is no first order or second order autocorrelation and the model fits the data well.

137
B-3: Regression output on sample excluding South Africa (with one lag of GDP)
Variable Regression #1 Regression #2

EMP(-1) -.0266064 (0.000)*** -.0263768 (0.000)***

MPR -4.247091 (0.062)* -4.25285 (0.052)**

PKFL -7.448417 (0.065)** -5.900472 (0.000)*

DEBT -1.631857 (0.000)*** -1.581775 (0.000)***

TOT - -215.3596 (0.000)***

CABA -4.435739 (0.013)** -

GDP 2.357813 (0.380) 4.591192 (0.381)**

GDP(-1) .5866153 (0.846) 2.881538 (0.381)

Wald Chi2, Prob. 1565.63 (0.0000) 11244.22 (0.0000)

Order 1 -1.3147 (0.1886) -1.3107 ( 0.1900)

Arellano-Bond Test
for Autocorrelation Order 2 -.04309 (0.9656) -.20786 (0.8353)

Sargan’s Test 10.56739 (1.0000) 9.605388 (1.0000)

***, ** and * denotes significance at 1%, 5% and 10% respectively. Values in parenthesis are the p-values. The
following diagnostic tests are run for the model: the Sargan test for over identification, the Arrelano-Bond test for
autocorrelation and the Wald Chi2 for model fit. The results indicate that the over identifying restrictions are valid,
there is no first order or second order autocorrelation and the model fits the data well.

138
B-4: Regression output on sample excluding South Africa (with 2 lags of GDP)
Variable Regression #1 Regression #2

EMP(-1) -.0320827 (0.000)*** -.0324843 (0.000)***

MPR -3.602594 (0.223) -1.024539 (0.746)

PKFL -2.870956 (0.704) .4274937 (0.929)

DEBT -1.352708 (0.034)** -2.209669 (0.003)***

TOT - -258.4076 (0.000)***

CABA -3.180687 (0.175) -

GDP 10.13556 (0.001) 7.30406 (0.097)*

GDP(-1) 7.452832 (0.116) 3.320865 (0.539)

GDP(-2) 11.45752 (0.000) 12.06022 (0.000)

Wald Chi2 531.95 (0.0000) 5813.25 (0.0000)

Order -1.3108 (0.1899) -1.312 (0.1895)

Arellano-Bond Test 1
for Autocorrelation Order -.38424 (0.7008) -.48361 (0.6287)
2
Sargan’s Test 11.7819 (1.0000) 9.584343 (1.0000)

***, ** and * denotes significance at 1%, 5% and 10% respectively. Values in parenthesis are the p-values. The
following diagnostic tests are run for the model: the Sargan test for over identification, the Arrelano-Bond test for
autocorrelation and the Wald Chi2 for model fit. The results indicate that the over identifying restrictions are valid,
there is no first order or second order autocorrelation and the model fits the data well.

139
Appendix B-5: Sample and data sources
Variable Measurement Source

Monetary Policy Rate Central Bank Discount Rate IFS

Terms of Trade Value of Exports to Imports IFS

Current Account Balance Current Account to GDP IFS

International Reserves Foreign Assets of Monetary IFS

Authority

Exchange rate Official Exchange Rate IFS

(per $)- period average

Public Debt Debt outstanding and World bank African

disbursed, (% of GDP) Development Indicators

Economic Growth Real GDP growth IFS

Private Capital Flows Private capital flows, total (% World bank African
of GDP) Development Indicators

Author’s Compilation

140