DEPARTMENT OF ECONOMICS UNIVERSITY OF NIGERIA, … Anthony Ph.D (Economics... · this thesis has...
Transcript of DEPARTMENT OF ECONOMICS UNIVERSITY OF NIGERIA, … Anthony Ph.D (Economics... · this thesis has...
i
Analysis of Financial Openness and Macroeconomic
Performance in Nigeria, 1986-2011
A Ph.D THESIS
BY
ORJI, ANTHONY
PG/PH.D/11/59630
DEPARTMENT OF ECONOMICS
UNIVERSITY OF NIGERIA, NSUKKA
Supervisor: Prof. C.C Agu.
ii
TITLE PAGE
Analysis of Financial Openness and Macroeconomic
Performance in Nigeria, 1986-2011
A Ph.D Thesis Report Presented to the Department of Economics,
University of Nigeria, Nsukka
In Partial Fulfillment for the award of Doctor of Philosophy (Ph.D)
in Economics
BY
Orji, Anthony
PG/Ph.D/11/59630
iii
APPROVAL PAGE
THIS THESIS HAS BEEN APPROVED BY THE DEPARTMENT OF
ECONOMICS, UNIVERSITY OF NIGERIA, NSUKKA FOR THE AWARD
OF Ph.D IN ECONOMICS
19/06/2014 19/06/2014
PROF. C. C. AGU DATE PROF. C. C. AGU DATE
SUPERVISOR HEAD OF DEPARTMENT
19/06/2014 19/06/2014
PROF. F. E. ONAH DATE PROF.AKPAN .H. EKPO DATE
INTERNAL EXAMINER EXTERNAL EXAMINER
19/06/2014
PROF. C. O. T. UGWU DATE
DEAN, FACULTY OF THE SOCIAL SCIENCES
iv
CERTIFICATION
This is to certify that the work embodied in this thesis is original and has not
been submitted in part or in full for any other degree of this university or any
other university.
19/06/2014 19/06/2014
PROF. C. C. AGU DATE PROF. C. C. AGU DATE
SUPERVISOR HEAD OF DEPARTMENT
19/06/2014 19/06/2014
PROF. F. E. ONAH DATE PROF.AKPAN .H. EKPO DATE
INTERNAL EXAMINER EXTERNAL EXAMINER
19/06/2014
PROF. C. O. T. UGWU DATE
DEAN, FACULTY OF THE SOCIAL SCIENCES
v
DEDICATION
To My Father and My God,
On whose wings of Love, Grace, Mercy and Favour, I am soaring and smiling.
And to all who have dreams & visions and believe they can accomplish them by His Grace.
vi
ACKNOWLEDGEMENTS
I sincerely appreciate the Lord God Almighty who started this journey for me and has also completed it. His love, grace, favour, mercy and wisdom have made the difference in my life. To Him be all the glory. I am grateful to my supervisor, Professor C.C Agu for his invaluable guidance and contributions which made the completion of this thesis a rewarding experience. May God keep on keeping you sir. My gratitude equally goes to my erudite Professors, fellow lecturers, non-academic staff and all students of the Department of Economics, UNN who stood by me, encouraged me and supported me in the course of my research work. Prof. F. E. Onah, Prof. N. I. Ikpeze, Assoc. Prof.Onyukwu E. Onyukwu, Prof. (Mrs.) S. I. Madueme, Dr. (Mrs.) Gladys Aneke, Rev. Fr. (Prof). H. E. Ichoku, Dr. F. O. Asogwa, Dr. P.C Ekeocha, Prof. Fonta, Dr. Moses O. Oduh, Dr. U. M. Ozughalu, Dr. I. A. Ifelunni, Dr. N. E. Urama, Dr Ugbor Kalu and Dr. J.I Amuka are greatly appreciated. I am also indebted to Profs. Jean-Yves Duclos, Luca Tiberti and other resource persons in the Dept. of Economics, University of Laval, Quebec Canada who provided me with a conducive environment to work as a Visiting Scholar during this period.
I want to thank in a very special way Prof. Akpan Ekpo, Dr. Chukwuma Agu, Dr. Emmanuel Nwosu, Dr. Jude Chukwu and Dr Ezebuilo Ukwueze for their guidiance, encouragements and support. I am also very grateful other colleagues of mine for having stood with me throughout this period: Dr. Jonathan E.Ogbuabor, Ilori Ayobami, Godstime Eigbiremolen, Vivian Nnetu, Uchechi Anaduaka. Johnson Ugwu, Chisom Emecheta, Mrs Ifeoma C. Mba, , Nelson Nkalu, , C. E. Nnaji, and Moses Nnaji. Men and women of God like Daddy G.O, Pst. Abraham O.Ugwa, Pst.Abraham Okorie, Rev. Dr. Dan Ozoko, Pst.Innocent Eleke,Pst.&Pst(Mrs) Favour Ochi, Dr.Parker Joshua, Pst. Bernard Agidi, Pst&Pst (Mrs) Wale Johnson, Pst.(Mrs)Lucy Obajaja, Pst. Goodness Kuzayet, Rev,Eze Ogbonnaya&Rev.Uche Okoji are deeply appreciated for their love and prayers.
I am eternally grateful to my lovely wife, my Angel and my Queen, Onyinye Imelda Anthony-Orji (Mrs), and my godly, blessed, favoured, and wise children for their candid love and support. I could remember that this programme started in 2011 shortly after our wedding and my Angel was very encouraging and patient with me even when I had to spend more time with my books and computer than “necessary”. Similarly, I thank my mother, Madam Ngozi Nnenna Orji and my brothers (and their wives) and sisters (and their husbands); Mr. Robert Orji (Papa Ejima), Grace Okezie (Mrs), Ebere, Chinyere Chima (Mrs) and other members of Pa Orji Okorie (of blessed memory) family. My in-laws, Mummy and Daddy Jim Ezeibe and their entire family are also acknowledged for their love, support and prayers. Other friends of mine; Dr.Goodluck Ebere Jonathan (Mr. President), President Obama, Peter Mba, Arinze Onyia, Chief Emeka Ani , Samuel Okereke and a host of others I cannot mention for lack of space are also appreciated. May God Almighty bless all of you in Jesus name, Amen.
Orji, Anthony (PG/Ph.D/11/59630)
vii
ABSTRACT The greatest challenge facing the country today is how to develop the economy and reduce poverty. Meeting this challenge is particularly difficult, if Nigeria should rely solely on domestic resources, given the low rate of savings and the attendant savings-investment gap. Against this background, following the Mackinnon and Shaw hypothesis, it becomes crucial to liberalize the financial system and also attract foreign resources into the economy. This thesis therefore, focuses on financial openness and macroeconomic performance in Nigeria. The impact of financial openness on economic growth in Nigeria is also investigated using quarterly data from 1986-2011. It equally examines the direction of causality between financial openness and economic growth as well as the impact of financial openness on output volatility in Nigeria. For empirical analysis, it uses two measures of financial openness: de facto (total capital flow) variables following Aizenman and Noy (2009) and de jure (Chin-Ito Index) based on Chinn and Ito (2012). The study applies the Autoregressive Distributed Lag Model based on unrestricted error correction model (ARDL-UECM), Granger Causality Test and the Generalized Autoregressive Conditional Heteroscedasticity (GARCH) Model to address the three objectives of the work. The results show positive impact of financial openness on economic growth in Nigeria both in the short run and in the long run. Interestingly, the de facto and de jure measures of financial openness is found to have similar degrees of impact on Economic Growth in the short run and long run respectively. The results also reveal that credit to the private sector is negatively associated with growth, indicating that there are problems with credit allocation and utilization in the country which could be occasioned by weak regulation/supervision and non-adherence to prudential guidelines in the financial system. The study also find that real interest rate has a positive relationship with economic growth. The results support the McKinnon-Shaw hypothesis, that is, in the long run interest rate liberalisation will ultimately lead to increased economic growth. Again, the thesis find the institutional quality variable contributing negatively and positively to growth in the short run and long run respectively. The second regression result indicates the existence of bidirectional causality between financial openness and economic growth in Nigeria. The findings of the thesis further show that none of the two measures of financial openness contributed to output volatility in Nigeria, within the period under review. From this work, our knowledge of the various measurement issues associated with financial openness has been enhanced and we can conclude that both measures are potent and robust for the Nigerian economy. Thus, the study recommends that government should continue to reform the domestic financial system while removing barriers to capital account transactions. And this should be done with every sense of objectivity, economic management dexterity and in line with global best practices. Furthermore, the country’s institutional quality should be comprehensively reviewed and upgraded. Strong emphasis should be placed on deepening the country’s democracy, reforming the governance and electoral systems, and reorganizing the socio/political structures of the country. Respect for the rule of law should be given priority by the leaders and the led. This is because according to our finding, poor governance, which is exemplified by corruption and lack of respect for the rule of law are detrimental to growth. However, if these anomalies are corrected, then sound financial openness policies and improved institutional quality will impact positively on growth in the long run.
viii
List of Tables
Table 1: Volatility and Average of Selected Macroeconomic Variables for diff. Economies 6
Table 2.1: Major Financial Sector Reforms in Nigeria from 1986 21
Table 2.2: Summary of Some Related Works on Financial Openness & Growth 28
Table 4.1: Summary of ADF Unit root test results of the series 64
Table 4.2: Bounds test for the estimation with De facto Financial Openness Variable 65
Table 4.3: Bounds test for the estimation with De Jure Financial Openness Variable 66
Table 4.4: The ARDL Model for the De facto Financial Openness Equation 66
Table 4.5: Parsimonious Restricted ARDL-ECM for De facto Financial Openness 69
Table 4.6: The ARDL Model for the De jure Financial Openness Equation 70
Table 4.7: Parsimonious Restricted ARDL-ECM for De jure Financial Openness 72
Table 4.8: Results of diagnostic tests 73
Table 4.9: The Granger Causality Test Results 80
Table 4.10a: Summary of ADF Unit root test results of the GARCH series 81
Table 4.10b: Descriptive characteristics of the variables 81
Table 4:11: Testing for Autocorrelation 82
Table 4.12: ARCH-LM Heteroskedasticity test 83
Table 4.13: The GARCH Model Results 83
Table 4.14: Arch test 85
List of Figures
Figure 1: GDP Growth Rate in Nigeria (1986-2011) 104
Figure 2: Real Interest Rate in Nigeria (1986-2011) 105
Figure 3: Inflation Rate in Nigeria (1986-2011) 105
Figure 4: Liberalization in the Neoclassical Growth Model 19
Figure 5. Cumulative Sum of Recursive Residuals (CUSUM) Test 79
ix
TABLE OF CONTENTS PAGE
Title Page ii
Approval Page iii
Certification iv
Dedication v
Acknowledgements vi
Abstract vii
List of Tables viii
List of Figures viii
Table of Contents ix
Chapter One
Introduction 1
1.1 Background to the Study 1
1.2 Statement of the Problem 4
1.3 Research Questions 8
1.4 Research Objectives 8
1.5 Research Hypotheses 9
1.6 Justification of the Study 9
1.7 Scope of the Study 11
Chapter Two
Literature Review 12
2.1 Conceptual Literature and Measurement Issues 12
2.2 Theoretical Literature 16
2.2.1 The Orthodox Theory 16
2.2.2 The Dependency Theory 17
x
2.2.3 The Neo-classical Counterrevolution Framework 18
2.2.4 Liberalization and the Neoclassical Growth Model 19
2.2.5 Major Financial Sector Reforms in Nigeria from 1986 21
2.2.6 Liberalization and Capital Flows in Nigeria 23
2.3 Empirical Literature 24
2.3.1 Financial Openness and Growth 24
2.3.2 Direction of Causality between Financial Openness and Growth 34
2.3.3 Financial Openness and Output Volatility 39
2.4 Shortcomings of Previous Studies 43
Chapter Three
Methodology 45
3.1 Theoretical Framework 45
3.2 Model Specification 47
3.3 Justification of the Models 54
3.4 Unit Root Tests 56
3.5 Cointegration Test and Estimation Procedure 56
3.5.1 The ARDL-UECM Procedure 56
3.5.2 The Granger Causality-ECM Procedure 58
3.6 The GARCH Model 59
3.6.3 Model Justification 61
3.6.4 Method of Estimation 62
3.7 Data Sources 63
3.8 Econometric Software 63
xi
Chapter Four
Presentation and Analysis of Results 64
4.1 Unit Root Tests 64
4.2 Bounds test 65
4.3 Estimation Results 66
4.4 Further interpretation and explanations of model parameters 74
4.5 Interpretation of the Parsimonious ARDL-ECM Models 75
4.6 ARDL- UECM and Short-run ARDL-ECM model diagnostic tests 78
4.7 Granger Causality Model Results 80
4.8 GARCH Model Results 80
4.9 Evaluation of Research Hypotheses 86
Chapter Five
Summary of Findings, Conclusions and Recommendations 87
5.1 Summary of Findings 87
5.2 Conclusion 88
5.3 Policy Implication of Findings and Policy Recommendations 90
5.4 Research Recommendations for Further Studies 93
References 94
Appendices 104
Appendix A: Graphs of Selected Macroeconomic Variables 104
Appendix B: The Normality Test 106
1
CHAPTER 1
INTRODUCTION
1.1 Background To The Study
Contemporary literature on economic development is replete with discussions on financial
openness and macroeconomic outcomes. This was sparked off by the seminal works of McKinnon
(1973) and Shaw (1973) which attributed financial repression as the cause of the unsatisfactory
growth performance of developing countries. Both McKinnon and Shaw advocated that financial
liberalization was needed to remedy the problems caused by the financial repressive policies of
developing countries. While this policy prescription initially generated some controversy, many
developing countries have adjusted their policies in the prescribed direction in recent years. In the
light of this, several countries, including developing and emerging economies have witnessed
some dramatic domestic financial/ capital account liberalization in the past three decades. The
opening of world economies and quest for greater integration also gave impetus for financial
liberalization and liberalization of the economies of both developing countries and emerging
economies. This is also in line with the “Washington Consensus”, which advocated for
liberalization of inflows, competitive exchange rate, interest rate liberalization, trade liberalization,
privatization, and deregulation of economic activities (Williamson, 1989; Lal, 2012).
Although, based on models of competitive and efficient markets, economic theory tells us that
financial openness should foster economic growth and development; empirical works so far have
not found indubitable evidence for the existence of such a link. While some countries have
benefited from financial liberalization, others have not enjoyed higher economic growth. Some
have even experienced some crises and recessions in the years following liberalization (Fratzscher
and Bussiere, 2004). Examples of this abound: Chile and Argentina in the early 1980s experienced
the negative effects of financial liberalisation. Mexico had their own negative experience between
1994 and 1995 and the Asian financial crisis equally affected many Asian Countries between 1997
and 1998, to name just a few. Also the global financial crisis of 2007–08 was triggered by, among
other things, insufficient financial market regulation (Bumann, et al, 2012). In their own view,
Andersen and Tarp (2003) equally argue that financial liberalisation in combination with a weak
regulatory structure may have strong adverse effects on growth.
2
The counter -argument to those underlining the benefits of openness based on the efficient- market
has been to stress the presence of market distortions that may lead to welfare- reducing effects of
financial openness. Such market distortions can take various forms, such as asymmetric
information and hidden action (Stiglitz, 1998) or be related to political economy factors (Bhagwati,
1998).
The literature on financial openness and growth has also been riddled with the controversy
regarding the direction of causality in their relationship. Financial openness is frequently presented
as beneficial for the economy or vice versa. However, regarding the direction and magnitude, there
is a controversy on the relationship (Aslanoğlu and Deniz, 2012). In line with financial
liberalization, a high level of capital tends to flow in the economies that are attractive for the
investors. Especially in emerging economies where charming elements are relatively higher than
elsewhere, foreign investors may direct their funds to these countries. However, financial openness
may easily lead to some alterations in the domestic system especially when external conditions
and effects that flow with liberalization are not properly checked. According to Bacchetta (1992)
it is likely that after financial liberalization, first, capital inflows will be observed. Together with
capital stock increase, domestic investments will be less and less charming as marginal
productivity declines. This decline ends up with capital outflows. He explains that higher domestic
interest rate lets foreign capital in and leads to appreciation first, but this is followed by capital
outflow due to arbitration in foreign and domestic interest rates which further leads to depreciation
of domestic currency; which consequently affects the economy. So the magnitude and direction of
causality between financial openness and economic growth may not be easily determined except
by empirical evidence.
While most of the empirical results linking financial openness and growth have been mixed and
conflicting, another fundamental issue that has recently caught the attention of researchers is the
relationship between financial openness and macroeconomic volatility. Some authors have argued
that financial openness could be a source of greater macroeconomic volatility, exposing vulnerable
countries to sudden reversals of capital flows (Stiglitz, 1998, Kaminsky and Reinhart, 1999).
According to this school of thought which emerged after the financial crisis of the 1980’s and
3
1990’s following capital account liberalization reforms by some countries, higher macroeconomic
volatility could be experienced by countries, either because they lack adequate financial
institutions to cope with large and sudden reversals of capital flows or because they lack policy
instruments to smooth cycles. In fact, Stiglitz (1998) suggested that the financial liberalization
thesis is “based on an ideological commitment to an idealized conception of markets that is
grounded neither in fact nor in economic theory”. On the other hand, some authors have argued
that opening the capital account can yield lower output volatility by promoting production base
diversification and enhance international capital flows (Razin and Rose, 1994, Bekaet, 2006).
From a welfare perspective there are two alternative ways to view the relationship between
financial openness and macroeconomic volatility. The first view is that financial openness should
help countries to untie consumption streams and output streams, allowing risk-averse agents to
smooth consumption and leaving output volatility inconsequential for welfare. Another way is to
consider that in addition to consumption volatility, output volatility is also detrimental to welfare.
In view of this, Ramey and Ramey (1995) are of the opinion that volatility has a detrimental impact
on output growth even after controlling for investment.
Following the global wind of liberalization, it is a common knowledge that Nigeria implemented
her Structural Adjustment Programme (SAP) in 1986. Before this period, interest rates in Nigeria
were generally fixed by the Central bank of Nigeria with periodic adjustments depending on the
government’s sectoral priorities (Agu, 1988; Uchendu, 1993). With the implementation of the
SAP, which focused on trade liberalization, the need for financial liberalization was also realized.
The steps that were taken in this regard were interest rate deregulation, introduction of an auction
market for treasury bills, identification of insolvent banks for restructuring, introduction of more
stringent prudential guidelines for banks, increase in banks’ minimum capital requirement and
upgrading and standardization of accounting procedures (Agu et al, 2014; Orji, et al 2014).
However, all of these measures were not implemented simultaneously. Interest rate deregulation
was the first step in 1986. Thereafter, the policy makers embarked on other major efforts of
financial liberalization. Legal reserve requirements were relaxed, credit controls were removed,
and the capital account was liberalized, among other measures.
4
1.2 Statement of the Problem
Prior to the introduction of Structural Adjustment Programme (SAP) in Nigeria in 1986, financial
liberalization had become an emerging trend in both developed and developing countries. This
was the prescription of the World Bank and the IMF following the structural imbalance and severe
economic woes experienced by developing countries as a result of oil price shocks, rigid exchange
and interest rate controls and escalating real interest rate for external debt servicing of the 1970s
and 1980s ( Agu et al, 2014; Okpara, 2010). Furthermore, the basic thrust of the economic reform
embodied in the SAP was deregulation, particularly, financial deregulation. Some developing
countries like Nigeria, Cameroun, Ghana, Botswana, Malawi, Senegal, Kenya and Zambia adopted
the liberalization of interest rate as a prominent feature of their financial reforms. Also interest
rates were fully deregulated in Indonesia, Philippines and Srilanka in the early 1980. While Nepal
freed most key interest rates in 1986, Korea, Malaysia and Thailand relaxed control by more
frequent advertisement (Fry, 1997).
The proponents of liberalization suggest that it is ideal for an economy. Honohan (2000) argues
that the process of financial liberalization is expected to increase the variability of interest rates
with its associated distributional consequences. The overall effect is to induce competition within
the financial services industry and in the entire economy, however, the experience of several
countries in the 1980s and 1990s indicate otherwise. For example Chile experienced some banking
problems right after deregulating the financial sector. Caprio and Kliengebiel (1995) also argue
that many banking systems experienced different problems after liberalization. Bakeart et al (2005)
suggest that in developing countries, financial liberalization may not yield intended benefits
because of the strength of domestic institutions and other factors. Demirguc-kunt and Detragiache
(1998) conclude that the benefits of financial liberalization should be weighed against the
increased potential for fragility.
As beautiful as the ‘message’ of liberalization sounds, there is a serious debate in the literature as
per whether its purported benefits are as real as projected by its proponents. For example, the
economic performance of Sub- Saharan African (SSA) countries, which have opened up their
capital accounts, has attracted considerable attention in recent years. The low rates of economic
growth and development experienced in these countries have from 1980’s to date, been described
5
as tragic. According to Babajide (2008), the average growth rate from 1961 to 2000 was 0.45%
for SSA, while it was 1.6% for Latin America and the Caribbean (LAC), 2.3 % for South Asia
(SA) and 4.9% for East Asia and the Pacific (EAP). For the Nigerian economy, the country has
experienced decades of slow development due to the unimpressive growth of her per capita income
and infrastructural deficits despite the liberalization of her financial system and capital account
(World Bank, 2012). Nigeria’s macroeconomic indicators have also been fluctuating since 1980.
For example, from early 1980s to the second half of the 1990s, annual inflation has averaged
around 30 percent. Subsequently, average inflation came down to one -digit rate. However, since
2001, inflation is back in the two-digit domain, with an average of about 12.50 %-22.17 % within
some years between 1986-2011 (NBS, 2012 and Akpan, 2013).
The causal relationships between Financial Openness and Growth have also created some concerns
among researchers in developed and developing economies like Nigeria. In developed countries,
several studies have tested for Granger causality between the two series using different samples
and estimation techniques but in Nigeria we have a dearth of such empirical findings, and where
it exists, it has been very inconclusive (Osinubi and Amaghionyeodiwe 2010, Oyatoye, et al 2011
and Ogbonna, et al, 2012).
Another key issue is the question of how financial openness impacts on output volatility. Some
empirical studies such as Greenaway, et al. (2002) and Serven and Schmidt-Hebbel (2002) have
also established that macroeconomic volatility can have adverse effect on growth and
development. The impact of financial openness on output volatility is very important for relatively
poor countries like Nigeria. Analyzing the performance of the Nigerian economy, World Bank
(2012) and CBN (2012) reveal that between 1986- 2012, broad macroeconomic aggregates in
Nigeria such as growth rate ( as shown in figure 1 in Appendix) , terms of trade , real interest rate
and inflation were among the most volatile in the developing world. For example the rate of growth
in Nigeria was 3% in 1986 when the liberalization process commenced but declined to -1% in
1987 and thereafter soared to 10% in 1988. It further declined to 7% in 1989 and rose again to 8%
in 1990. From 1990 to 1994, there was a continuous decline reaching as low as 1.3% in 1994.
There was a little improvement between 1995 and 1996 when the economy grew from 2% to 4%,
but it began to decline again from 3% in 1997 to 1% in 1999, the very year Nigeria embraced a
6
new democratic leadership. Owing to some policy initiatives, the economy improved to 5% in
2000 but declined again to 3% and 2% in 2001 and 2002 respectively. However, when the new
government came on board in 2003, the economy grew again to 10% and 11% in 2004. The growth
was short-lived as the economy nosedived again to 5% in 2005 and increased minimally to 6%
between 2006 and 2008.The economy experienced another growth between 2009 and 2010 at 7%
and 8% respectively but declined again to 7% in 2011. These fluctuations and volatility in growth
rates of the economy have been attributed to many factors ranging from economic mismanagement
to erratic policy reversals. To buttress this point further, the table below shows the relative
performance of some macroeconomic variables in Nigeria when compared with other selected
emerging and developed economies.
Table 1: Volatility and Average of Selected Macro economic Variables for different Economies
Countries GDP Growth Rate Inflation Exchange Rate
Developed Economies Volatility Average Volatility Average Volatility Average
Canada 2.04 2.51 1.41 2.50 0.16 1.29
South Korea 4.01 6.40 2.14 4.40 476.46 631.33
Emerging Economies
Indonesia 4.08 4.99 10.47 10.52 3,757.88 5,780.65
South Africa 2.19 2.48 4.41 9.19 2.46 5.31
Egypt 1.74 4.46 7.03 11.02 1.76 3.69
Nigeria 6.23 4.61 18.58 22.17 57.32 62.90
Source: World Bank (2012) and Author’s Computation based on 5-year Averages (Means) and Standard Deviations of the selected macroeconomic variables for the various Economies (1986 – 2011)
Table 1 above shows the averages and volatilities of selected macroeconomic variables in some
selected economies. On average, the volatilities of Gross Domestic Product (GDP) growth,
inflation, and exchange rates are higher in emerging market economies than in developed
economies. Among emerging market economies, Nigeria exhibits the highest GDP volatility.
7
Besides, average inflation is higher in emerging market economies relative to developed
economies. Nigeria’s average inflation over this period is the highest in the emerging market
economies group, followed by Egypt and Indonesia. Inflation rate variability in Nigeria is also
high, although with slightly smaller volatility than that of Indonesia. Furthermore, exchange rate
variability in Nigeria is higher than Canada, South Africa, and Egypt but lower than South Korea
and Indonesia. On the other hand, Nigeria’s GDP growth showed some buoyancy among the
selected developed and emerging market economies, second only to South Korea and Indonesia,
which all grew on average around 5 to 6 percent, a year, within the period under review. Although,
Nigeria’s GDP showed some increase within the period, yet it is the most volatile among the
countries compared. This shows that Nigeria faces a macroeconomic environment that is indeed
more volatile than say Canada, South Korea, Indonesia, South Africa, and Egypt, at least in terms
of GDP Growth, inflation and exchange rates. This development further corroborates the finding
of Batini (2004) that “emerging market economies (like Nigeria) face more volatile
macroeconomic environment, and typically have weaker institutions that enjoy less credibility than
their developed economies counterparts”.
Over the last three decades, high macroeconomic volatility has become a key determinant as well
as consequence of poor economic management. This is in line with Kama (2006), which posits
that the ability of the financial sub-sector to play its role has been periodically punctuated by its
vulnerability to systemic distress and macroeconomic volatility. The Nigerian economy has also
been characterized by low growth trap as a result of low savings-investment equilibrium. With an
average annual investment rate of about 16% of GDP, Nigeria is still far behind the minimum
investment rate of about 30% of GDP required to minimize poverty and stimulate real growth
(CBN, 2010, World Bank, 2010).Real Interest Rate Movement and Inflation rate have also being
volatile (as depicted by figures 2 and 3 in Appendix). In addition to this, fiscal policy in Nigeria
has been characterized by highly volatile, inefficient and unsustainable public sector spending.
Another key issue here is the question of how to measure financial openness. Two broad
approaches can be found in the literature: one based on measuring de jure openness and the other
measuring de facto openness {(Raddatz, (2007); Fratcher and Bussierre, (2004); Lane and
Millessi-Ferreti 2005; Edison et al 2002b and Kray (1998)}. De jure openness is often proxied by
8
the removal of restrictions to capital account transactions as published in the IMF’s Annual Report
on Exchange Arrangements and Exchange Restrictions (AREAR). For the de facto openness
measures, different studies have used different capital flow variables. Each of these measurements
when adopted for cross-country regressions have their pitfalls respectively. Thus, these problems
call for country-specific regressions. Also despite the efforts to promote the ideals of domestic
financial market cum capital account openness in Nigeria through competitive market framework,
there is still the fundamental challenge of understanding its real impact on the economy. Thus,
macroeconomic outcomes resulting from financial openness in Nigeria is still largely unexplored.
Hence, further empirical investigation is needed to unravel the outcomes of financial openness
policies with respect to the growth and volatility of the Nigerian economy, using the de facto and
de jure approaches.
1.3. Research Questions
Based on the above discussions, this work intends to address the following research questions:
(1) What is the impact of financial openness on economic growth in Nigeria?
(2) Is there any evidence of the existence of a causal relationship between financial openness and
economic growth in Nigeria?
(3) Does financial openness contribute to output volatility in Nigeria?
1.4. Research Objectives
The overall objective of this study is to estimate and critically analyze the relationship between
financial openness policies and the performance of the Nigerian economy using the de facto and
de jure measurement approaches. Specifically, this work intends to achieve the following
objectives:
(1) To examine the impact of financial openness on economic growth in Nigeria.
(2) To investigate the existence of a causal relationship between financial openness and economic
growth in Nigeria.
(3) To determine the impact of financial openness on output volatility in Nigeria.
9
1.5. Research Hypotheses
The hypotheses to guide this research work include:
Ho1: Financial Openness has no significant impact on Economic Growth in Nigeria
Ho2: There is no significant direction of causality between financial openness and economic
growth in Nigeria
Ho3: Financial Openness has made significant impact on output volatility in Nigeria
1.6. Justification of the Study
The Liberalization thesis has been one of the key issues at the fore front of development
macroeconomics and finance as the influence of the epoch-making studies of McKinnon (1973)
and Shaw (1973) has spread. It is pertinent to note that according to Fanelli and Medhora (1998),
the results of liberalization policies in different countries have been mixed:
(i) Although there is an increase in credit supply after liberalization, the result of financial
deepening is rather modest;
(ii) There have been open financial crises a few years after financial liberalization was adopted;
(iii) Some interventionist countries like South Korea have achieved impressive levels of financial
deepening and growth without significant liberalization.
Therefore, as the debate on the benefits of financial liberalization policy goes on in the literature,
this work will help policy makers to gain a deeper understanding of how this policy prescription
has contributed (or not contributed) to economic growth and output volatility in Nigeria. It may
look like a logical and inescapable step for an emerging economy like Nigeria, with ambition to
optimize its interaction and engagement with the international financial markets, but this study
will reveal some empirical truths that will guide the government and policy makers towards
implementing and maximizing the liberalization policies. Indeed the findings will guide policy
10
makers in designing and implementing macroeconomic and financial policies that will enhance
national development and improve the social welfare of the populace.
Also this work will enhance our knowledge of the various measurement issues associated with
financial openness and guide us in identifying the ones that are potent and/or the ones that are less
efficient. The first leads to the issue of whether, and how government should remove barriers to
capital account transactions. The second question requires policy-makers to decide, given that
legal barriers have been removed, how best to manage capital flows. Also, this research work will
analyse the role of both de facto and de jure openness measures and this will assist greatly in
unraveling the measurement technique that is most robust for the Nigerian Economy. Estimating
the causal relationship between financial openness and economic growth would also be
informative in predicting how economic growth will be affected if policymakers are to change
financial openness policies and vice versa.
In addition to this, as we pursue the Millennium Development Goals, Financial System Strategy
(FSS) 2020, and Vision 20:2020 which are basically aimed at poverty reduction, enhancement of
peoples’ welfare and quality of life, making Nigeria the safest and fastest financial hub of Africa,
engendering financial inclusion and stability and finally positioning Nigeria in the league of the
world’s top 20 economies by the year 2020, this study will elucidate on how financial openness
policies have impacted on the stability and volatility of the macro economy. This is because what
happens at the macro economy has direct and indirect impacts on the lives of the entire citizens.
Furthermore, we shall also utilize some macro models to address our specified objectives, bearing
in mind that Nigeria is still a developing country with several constraints. This is essential because
some methodologies that were adopted by other authors to study developed economies or
undertake cross-country studies may not adequately fit into our own domestic needs in Nigeria.
It will also be interesting to note that financial openness is a fascinating topic to study for
researchers of development macroeconomics and finance not only because of its compelling policy
relevance but also because of the enormous variations of approaches and experiences across
countries. Differences in speed and approach to liberalization have often been driven as much by
philosophy, political circumstances, and regional fads as by economic factors. Hence, a country-
11
specific study of the effects of financial openness can potentially reveal a wide array of natural
variation in experiences.
To the best of my knowledge, this thesis is among the first to examine in detail, the dynamic impact
of financial openness on overall economic growth and volatility in Nigeria using Capital flow
variables, the Chinn-Ito index, the ARDL – Bounds testing approach and GARCH Model.
Finally, this thesis will serve as a good reference material to other researchers who wish to explore
more empirical issues following our subject of discussion.
1.7. Scope of the Study
Our study spans through the period 1986- 2011 to account for the commencement of the
liberalization exercise in Nigeria. As we noted earlier, Nigeria implemented her Structural
Adjustment Programme (SAP) in 1986. Before this period, interest rates in Nigeria were generally
fixed by the Central bank of Nigeria with periodic adjustments depending on the government’s
sectoral priorities. With the implementation of the SAP, which started with trade liberalization, the
need for financial liberalization was also realized. Consequently, the capital account was
liberalized and opened up to give room for free flow of capital across our national borders. In this
study, our variables of interest for econometric analysis shall be de facto (capital flows) variables,
which are the sum of FDI, portfolio flows and other investments following Aizenman (2004 and
2008), and Aizenman and Noy (2009) and de jure (Chin-Ito Index) based on Chinn and Ito (2012).
12
CHAPTER 2
LITERATURE REVIEW
2.1 Conceptual Framework and Measurement Issues
A country’s balance of payment has two major accounts namely current and capital account. The
Current account details economic transactions which provide income for the recipient country.
These transactions include trade in goods (visibles), trade in services (invisibles), payment of
factor income (dividends, interests and migrant remittances from earnings abroad), and
international transfers (gifts).Capital account on the other hand represents a variety of financial
flows like foreign direct investment, portfolio/ equity flows (or portfolio/equity investments),
loans, acquisition of assets in one country by foreign residents, etc. Capital account openness
therefore refers to a process whereby there is a systematic reduction or removal of restrictions on
capital flows to a country. Also, it connotes a deliberate policy that allows domestic businesses to
borrow from foreign banks, and foreigners are allowed to purchase domestic debt instruments as
well as invest in the domestic stock market (Henry and Lombard, 2003).
Furthermore, following Agosin and Mayer (2000); Levine, et al (2002), Agenor (2003), and
Carnignani and Chowdhury (2005), financial liberalization may involve the process of liberalizing
domestic financial markets and lifting administrative or legal restrictions on capital movements;
and hence creating the necessary conditions for the integration of the domestic financial system
into the global market. In this study, however, we use capital account openness/financial openness
interchangeably. A key component of the argument here following Kose et al, (2008) is that it is
not just the capital inflows themselves, but what comes along with the capital inflows, that drives
the benefits of financial openness for a developing country like Nigeria. There is considerable
evidence that financial openness if properly managed, serves as a catalyst for a number of benefits.
Some of these benefits may be termed “collateral benefits” if they are not the primary motivations
for a country to undertake financial openness. These collateral benefits could include development
of the domestic financial sector, improvements in institutions (defined broadly to include
governance, rule of law, etc), better macroeconomic policies, etc. These benefits then result in
13
higher economic growth, usually, through gains in allocative efficiency. The diagram below
represents the two view points on how financial openness could interact with growth and volatility.
The Traditional View
GDP Growth
Financial Openness GDP Growth
Volatility
The traditional view focuses on the importance of channels through which financial openness and
capital flows could increase GDP Growth and reduce volatility.
An Alternative Perspective
Financial Openness GDP Growth
Volatility
More efficient international allocation
of capital
Capital deepening
International risk-sharing
Potential Collateral Benefits
Financial Market Development
Better governance
Institutional Development
Traditional Channels
Macroeconomic Discipline
14
This alternative perspective based on Kose et al (2008) acknowledges the relevance of the
traditional channels, but argues that the role of financial openness as a catalyst for certain
“collateral” benefits may be more important in increasing GDP growth and reducing volatility.
Another key conceptual issue in this study is the question of how to measure financial openness.
Two broad approaches can be found in the literature: one based on measuring de jure openness
and one measuring de facto openness. De jure openness is mostly proxied by the removal of
restrictions to capital account transactions as published in the IMF’s Annual Report on Exchange
Arrangements and Exchange Restrictions (AREAER). For de facto openness, Edison et al (2002b)
and Kraay (1998), for example, used seven variables- four based on FDI and portfolio flows
(combined FDI and portfolio net flows, combined FDI and portfolio inflows, FDI inflows,
portfolio inflows) - two proxies related to the size and composition of foreign debt (total foreign
debt and short-term foreign debt and trade openness –defined as the sum of exports and imports).
Moreover, they employed two proxies for stock variables (combined FDI and portfolio net stocks,
combined FDI and portfolio in-stocks). Net flows and stocks refer to the difference between the
asset and liabilities sides of the balance of payments (B.O.P) in a particular period. Aizenman and
Noy (2009), Ozdemir and Erbil (2008), and Kose et al (2008) equally adopted various capital flow
variables in the literature, to study de facto financial openness of different economies. The
AREAER measure has also been utilized in different forms in the literature. The usual way is to
simply define it as a discrete 0-1 variable, that is, indicating full ‘openness’ or ‘closedness’. Studies
using longer time periods, such as 5-year periods, generally use the share of the years in which a
country had an open capital account as the measure of openness. The advantage of these measures
is that they allow for a clear and easy identification of when a country had removed all barriers to
capital account transactions. However, a drawback is that countries may liberalize their capital
accounts by removing individual barriers gradually over time.
As an alternative, Quinn (1997) exploits the details of the description in the AREAER to construct
an openness measure which can take 9 different degrees of openness – from 0 to 4 in 0.5 point
increments. This allows for a much finer categorization of de jure openness and its changes.
However, a key drawback is that this openness measure has been created only for four years- 1958,
15
1973, 1982 and 1988- thus, not allowing the identification of those years in which a country
undertook those changes.
In this study, we adopt the capital account openness index developed by Chinn and Ito. The
KAOPEN index was used by Chinn and Ito (2002, 2006 and 2012) in their studies of the
determinants of financial development. The researchers found that the rate of financial
development, as measured by private credit creation and stock market activity, is linked to the
existence of capital controls, and that higher level of financial openness contributes to the
development of equity markets if a threshold level of institutions is attained, which is more
prevalent among emerging market countries. This index was also updated by Chinn and Ito (2012)
and it is available for use by researchers conducting empirical analysis in this area of research. The
construction of the KAOPEN index is based on the first principle component of four binary
variables in IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions
(AREARER) and it takes higher values for more open financial regimes. These four variables are
defined as follows: K1 is the variable that indicates the presence of multiple exchange rates; K2 is
the variable that indicates restrictions on current account transactions; K3 is the variable that
indicates the restrictions on capital account transactions; and K4 is the variable that indicates
requirements of the surrender of export proceeds. One important merit of the index is its wide
coverage; it is available for more than 182 countries and for a long time period (1970 through
2011)
As mentioned by a number of authors (such as Edison et al., 2002), it is not easy to measure the
extent of openness in capital account transactions. By nature of its construction, the KAOPEN
index is considered to be a de jure measure of financial openness because it attempts to account
for regulatory restrictions on capital account transactions. Hence, this index is different from price-
based measures of financial openness, often referred to as de facto measures of financial openness.
One might argue that a de jure measure is a poor indicator of openness in the sense that releasing
controls do not necessarily lead to more cross-border transactions. The response to such concerns
is that the factors determining the magnitude of capital flows are many. The investment climate in
the country, as well as the culture might influence capital inflows. The policy tool that is most
directly related to the regulation of capital account transactions is capital account liberalization, i.e
16
eliminating the barriers to allow access. Whether a change in the rules helps increase the magnitude
of capital flows is another question.
As an alternative to these de jure measures, the literature has also analysed various de facto proxies
of openness. The rationale for looking at actual openness is that a country that is open de jure may
not necessarily experience such inflows. Since the question of interest is whether capital flows
benefit or hurt countries, one may argue that one should define openness in terms of both legal
restrictions (de jure) and actual capital flows (de facto). The literature has looked at various capital
flows related to FDI, portfolio flows and other investment flows (e.g. Aizenman 2008). Thus, in
this study we adopt Aizenman and Noy (2009) capital flow variables to measure our de facto
financial openness.
2.2 Theoretical Literature
Theoretically, financial openness is expected to engender flow of resources from capital- surplus
industrial countries to capital- deficit developing and emerging countries. However, there have
been some serious theoretical debates on the issues of liberalizing capital across national borders
with different schools having different views of international mobility of capital. These thoughts
are tailored mainly along the Orthodox, Dependency, and Neoclassical-counter-revolution
frameworks (Mailafiah, 2006).
2.2.1 The Orthodox Theory
Liberalization is seen by main stream economists as a means of solving global economic problems
definable in terms of wants, exchange, global resources, and growth. This model views capital
mobility as adding new resources, technology, management and competition to capital deficit
economies in a way that improves efficiency and stimulates change in a positive direction.
Currently the example of the Asia Tigers is used to drive home the growth-driven force of capital
mobility when FDI flows are encouraged with liberalization of capital account transactions. This
submission transcends the classical and neo-classical viewpoints.
Neo-classical theory suggests that free flows of external capital should equilibrate and smoothen
a country’s consumption or production path. In the real world, this theory has been disputed
because liberalization of the short term capital account has been associated with serious economic
17
and financial crises in Asia and Latin America in the 1990s which has necessitated the caution in
the 21st century to fully liberalize the capital account transactions. Long term flows are regarded
as much more stable and there is the suggestion that developing countries may wish to liberalize
only long –term flows while still controlling partially or wholly short-term flows. These
viewpoints have been contentious within the framework of globalization and pressure for full
integration of world financial markets.
Capital account openness has been identified as a necessary strategy to attract private capital flows
to substitute declining aids in developing countries {(Grill and Milesi-Ferrti, (1995), Quinn (1997),
and Demirguc-Kunt and Detragiache (1998)}. In these studies, it is imperative to note that capital
account liberalization correlated with growth as well as the deepening of the financial sector.
2.2.2 The Dependency Theory
This school of thought is tailored along the neo-Marxist analysis which developed from Marxism.
This view could be summarized as dependence on capital-surplus developed economies by the
capital deficit developing economies. Though unpopular as a result of the collapse of communism
in the 1980s and the subsequent acceptance of the market doctrine by the former Eastern bloc, it
helps historically to examine the diverging viewpoints of development economists.
According to this school which was popularized by Frank (1975), it is assumed that the dependence
worsens the conditions of the developing countries and engenders underdevelopment. Thus, the
penetration of capital from developed countries into developing nations through FDI flows and
short term capital cannot produce beneficial results in the host countries. The assumption is that
there exists a symbiotic relationship between the metropolis (developed) countries and the
underdevelopment of the satellite (developing) countries and that capital mobility to the
developing (satellite) nations is mainly to benefit the metropolis.
The structuralist import-substituting and capitalist industrialization strategy in Latin America in
which the “Foreign Monopoly Capital” was taking over the import substitution process was further
analyzed by Frank (1975). Frank noted that the strategy was unprogressive and that the peripheral
formations became more underdeveloped with their incorporation into the world capitalist system.
The theorists recommended the need to severe link with the exploitative international capitalism
18
as the recipe to developing the economies of the periphery. Revolutionary as this may sound; it is
unattainable in a world that is fast becoming a global village.
In his own view, Aremu (2005) posits that a modification of this thought has been formulated
drawing from the experiences of the newly industrializing economies (NIEs) of Latin America and
South East Asia. In these countries, foreign investors were attracted through the provision of
enabling environment, while other modalities for entry and operations were negotiated. Aremu
(2005) also suggest that the modification of this model presupposes that through a strategy of
autonomous and self-reliant macroeconomic policy objectives and implementation programmes,
developing nations can still use external stimuli, particularly FDI to achieve their development
aspirations and strategies.
2.2.3 The Neo-classical Counterrevolution Framework
This school of thought emerged following the questioning of the relevance of the dependency
argument at the end of the 1970s. The neoclassical counter revolution was therefore launched with
a re-affirmation of the dictates of the market and the importance of “getting the prices right”
(Mailafia, 1997). The counterrevolution, led by Ian Little, Bela Ballasa, Anne Krueger and Deepak
Lal, argued that the policy-induced distortions of developing countries largely responsible for their
poor development performance, and proposed that the problems of economic development can
only be solved by an economic system with freely operating markets and a minimalist government
(Ohiorhenuan, 2003). This formed the theoretical underpinnings of the structural adjustment
programmes of the 1980s.
The World Bank publication, “Accelerated Development in Sub-Saharan Africa: An Agenda for
Action” (World Bank, 1981) emphasized the importance of correct pricing policies and reduced
government intervention in economic activities as the two main keys to a revival in African growth
rates. Thus, the IMF conditions for access to her facilities included not only control of money
supply, but removal of price distortions including price controls , subsidies, foreign exchange,
tariffs, freeing of the markets from public sector intervention and elimination of restrictions against
foreign direct investment. An outcome of the protest against the harsh conditions of the IMF policy
prescriptions was the emergence of the “Washington Consensus” emanating from the IMF, World
19
Bank and the group of seven leading industrial countries, particularly United States of America.
The consensus advocated a focus on balanced budget, liberalization of trade and financial flows,
exchange rate correction, domestic market deregulation and privatization.
2.2.4 Liberalization and the Neoclassical Growth Model
This section illustrates the fundamental predictions of the neoclassical growth models about the
impact of capital account liberalization on a developing country. Following Diercks (2012) and
Henry (2006) who have also utilized this framework based on Solow, (1956) and Swan (1956), we
submit that understanding this framework can also be useful in the context of a developing country
like Nigeria, which is the focus of our study. The diagram below elucidates this point further.
However, our theoretical framework and model building in chapter three of this thesis follow the
Endogenous Growth Model.
Output per Unit of
Effective Labor (n + g +δ) k
sf (k)
A
ks.state k*s.state K
Figure 4: Liberalization in the Neoclassical Growth Model
Assume that output is produced using capital, labor, in a Cobb-Douglas Production
Function with labor-augmenting technological progress:
Y=F (K, AL) =Kα (AL) 1-α (2.1)
Let 푘 =be the amount of capital per unit of effective labour and 푦 =
the amount
20
of output per unit of effective labor. Using this notation and the homogeneity of the production
function we have:
y=f (k) =kα (2.2)
Let s denote the fraction of national income that is saved each period and assume that capital
depreciates at the rate δ; the labor force grows at the rate n, and total factor productivity grows at
the rate g. Savings each period builds up the national capital stock and helps to make capital more
abundant. Depreciation, a growing population, and rising total factor productivity, all work in the
other direction making capital less abundant. The following equation summarizes the net effect of
all these forces on the evolution of capital per unit of effective labor:
k (t) =sf (k (t)) - (n +g+δ) k (t) (2.3)
When k (t) = 0, the economy is in the steady state depicted by Point A in Figure 4 above. The ratio
of capital to effective labor (k) is constant at Point A. In contrast, the steady state level of capital
(K) is not constant, but growing at the rate n + g. Output per worker
grows at the rate g.
Finally, the steady state marginal product of capital equals the interest rate plus the depreciation
rate:
f'(ks.state)=r+δ (2.4)
Equation (2.4) gives a general expression of the equilibrium condition for investment. This
equation has important implications for the dynamics of a country’s investment and growth in the
aftermath of capital account liberalization, because the impact of liberalization works through the
cost of capital. Let r* denote the exogenously given world interest rate. The standard assumption
in the literature is that r* is less than r, because the rest of the world has more capital per unit of
effective labor than the developing countries. It is also standard to assume that the developing
country is small, which means that nothing it does affects world prices (Henry, 2006).
Under these assumptions, when the developing country liberalizes, capital surges in to exploit the
difference between the world interest rate and the country’s rate of return to capital. The absence
of any frictions in the model means that the country’s ratio of capital to effective labor jumps
immediately to its post-liberalization, steady state level. Figure 4 depicts this jump as a rightward
shift of the vertical line from ks.state to k*s.state.
21
In the post-liberalization steady state, the marginal product of capital is equal to the world interest
rate plus the rate of depreciation:
f'(k*s.state)=r*+δ (2.5).
According to Barro and Sala-i-Martin (1995), the instantaneous jump to a new steady state implies
that the country installs capital with speed. Although this may not be completely true, however,
the vital fact about the transition dynamics is that there must be a period of time during which the
capital stock grows faster than it does before or after the transition. To see why the growth rate of
the capital stock must increase temporarily, recall that in the pre-liberalization steady state the ratio
of capital to effective labor (ks.state) is constant, and the stock of capital (K) grows at the rate n + g.
In the post-liberalization steady state, the ratio of capital to effective labor (k*s.state) is also constant
and the capital stock once again grows at the rate n + g. However, because k*s.state > ks.state, it follows
that at some point during the transition, the growth rate of K must exceed n + g. (Henry, 2006 and
Diercks, 2012).
2.2.5 MAJOR FINANCIAL SECTOR REFORMS IN NIGERIA FROM 1986
Table 2.1 MAJOR REFORMS YEAR
INTEREST RATES LIBERALISATION
Two foreign exchange markets established.
Interest rate controls completely removed. Bank licensing liberalized. Foreign exchange market unified.
Foreign exchange bureaus established. Bank portfolio restrictions relaxed.
Banks permitted to pay interest on demand deposits. Auction markets for government securities introduced. Capital adequacy standards reviewed upwards. Extension of credit based on foreign exchange deposit banned.
1986
1987
1987
1988
1988
1989
1989
1989
1989
22
Risk-weighted capital standard introduced and banks’ required paid- Capital. Uniform accounting standards introduced for banks. Stabilization securities to mop up excess liquidity introduced. Bank licensing emerged. Central bank empowered to regulate and supervise all financial institutions Interest rates re-administered. Interest rate control removed once again. Privatisation of government owned banks begun again. Capital market deregulation commenced. Foreign exchange market reorganized. Credit control dismantled. Indirect monetary instruments introduced. Five banks taken over for restructuring. Interest and exchange rate controls re-imposed.
1990
1990
1990
1991
1991
1991
1992
1992
1992
1993
1993
1994
SECURITIES MARKET LIBERALISATION
Liberalization of capital/equity flows. Continuation of interest controls initiated fiscal reforms. Exchange controls relaxed. Autonomous foreign exchange market
introduced.
Liberalization of capital market continues. Retention of interest controls continuation of fiscal reforms. Official foreign exchange market operation by government transactions continued operation of the autonomous foreign exchange market.
1995
1995
1996
1996
1996
INCREASE IN BANK & CAPITAL MARKET RECQUIREMENTS
Recapitalization of banks and stock market. Five banks taken over by the CBN. Assets Management Company of Nigeria established.
2005 2009 2010
Source: Author’s compilation based on Ikhide and Alawole (2001), Agu et al (2014), Orji et al (2014) and various CBN publications.
23
2.2.6 Liberalization and Capital Flows in Nigeria
The postulation of the McKinnon-Shaw hypothesis emphasizes that financial liberalization will
lead to an increase in savings, investment and subsequently, rapid economic growth. The Structural
Adjustment Programme (SAP) which was embraced by Nigeria in 1986 marked the beginning of
the liberalization of the Nigerian economy. Prior to the period, the economy was regulated, and
that affected the free movement of capital necessary for economic growth. One of the outcomes of
the financial liberalization as identified in the literature is the increase in capital flows and the
consequent augmentation of the country’s foreign exchange reserves (Owusu, 2012 and Agu et al,
2014). Furthermore, foreign direct investment (FDI) flow into the country increased from an
average of 1.3 % GDP (pre-liberalization) to an average of 2.6 % of GDP after liberalization. The
increase in FDI could be attributed to the increase of foreign investment in the oil and
communication sectors of the economy. At the commencement of the Structural Adjustment
Policy, the stock of total reserves (excluding gold) in Nigeria was $1.165 m in 1987. This figure
increased to $53.002m by the end of 2008. This increase in capital flows and reserves could also
be attributed to flexible exchange rate regime adopted after the adoption of Structural Adjustment
Programme (SAP) in 1986. Furthermore, the SAP heralded a lot of policy reforms that led to the
publication of an Industrial Policy for Nigeria in 1989. Critical policy reforms leading to the
changes in the investment climate in Nigeria for both domestic and foreign investors include, the
abrogation of the Exchange Control Act of 1962 as well as Nigerian Enterprises Promotion Decree
1989 and their subsequent replacements with the Nigerian Investment Promotion Council Decree
No 16 of 1995 and Foreign Exchange (Monitoring and Miscellaneous Provisions) Decree 17 of
1995. As it were, the country did not record any substantial Net Portfolio Investment (NPI) on her
Balance of Payment (BOP) until 1986. However, between July 1995 and July 1996, about US$6.0
million foreign portfolio investment (FPI) was made in the Nigerian capital market through the
Nigerian Stock Exchange (NSE) for the first time since 1962, while for the whole of 1996, foreign
investment through the Nigerian Stock Exchange totaled UD$32.99 million {(Onosode (1997) as
cited in Obiechina (2010)}.
24
2.3 Empirical Literature
2.3.1 Financial Openness and Growth
Despite intensive research devoted to this issue, the literature on the effects of financial openness
on growth has produced conflicting, and sometimes, contradictory results (Eichengreen, 2001,
Mishkin, 2007).
Building on the work of Schumpeter (1911), financial liberalization as advocated by McKinnon
(1973) and Shaw (1973) is a deliberate attempt by developing countries to move away from
financial repression. The models of McKinnon (1973) and Shaw (1973) introduce financial
development as a process and strategy to achieve faster economic growth. They find that
liberalization from restrictions such as interest rate ceilings, high reserve requirements, and
selective credit programme, facilitates economic development. In addition, they argue that positive
real interest rates lead to more efficient credit allocation which provides an additional impact on
growth.
Quinn (1997) was one of the first studies to identify a positive relationship between capital account
liberalization and growth. Quinn’s empirical estimates find that the change in his measure of
restrictions on capital account liberalization has a strongly significant effect on the growth in real
GDP per capita in his cross section of 58 countries over the period 1960- 1989.
Studies by French and Poterba ( 1991), Tesar and Werner (1995), Baxter and Jermann (1997) and
Lewis (1999), argue that financial account openness stimulates capital accumulation , productivity
growth and economic growth by relaxing financial constraints through greater access to external
capital, by promoting more disciplined macroeconomic policies under international pressure,
enhancing production specialization through risk-sharing and increasing the functioning of
domestic financial systems through the importation of financial services and intensification of
competition.
Klein and Olivei (1999) in a study of a cross section of 82 industrial and non-industrial nations
find a positive effect of capital account liberalization on growth among industrial countries, but
they do not find evidence that capital account liberalization promotes growth in non-industrial
countries. This significant result seems to be because of the presence of the OECD countries in the
25
sample. Klein and Olivei show that capital account liberalization significantly affects the change
in financial depth in a sample consisting of 20 OECD countries but not in a sample of the non-
OECD countries, nor in a narrower non-OECD sample of 18 Latin American countries known to
have had a relatively high incidence of capital account liberalizations. They also estimate a growth
model that includes the change in financial depth as regressor and find that financial development
is a significant determinant of growth per capita. They conclude that the beneficial effects of capital
account liberalization, at least with respect to promoting financial depth, are achieved only in an
environment where there is a constellation of other institutions that can usefully support the
changes brought about by the free flow of capital. Billiu (2000) also finds that capital account
liberalization spurs growth by promoting financial development.
Henry (2004) argues that if a developing country opens its stock market to foreign investors,
aggregate dividend yield falls by 240 basis points, growth rate of output increases by an average
of 1.1 percentage points per year, and the growth rate output per worker rises by 2.3 percentage
points per year. Also, Bekaert, et al (2005) show that foreign investors pressure local institutions
to adhere to international standards in order to improve local corporate governance and reduce the
division between internal and external finance.
Loyayza and Ranciere (2006) provide the summary of the relationship between financial
liberalization and economic growth. Their modern growth model reveals that the financial sector
impacts capital accumulation as well as the rate of technological development. Serven (2002) also
observes that financial openness grants markets dominant role in setting financial asset prices and
returns, allocating credit, and developing a wider array of financial instruments and intermediaries.
Fratzscher and Bussierre (2004) analyse the openness-growth nexus for a set of 45 developed
countries and emerging market economies: 11 OECD, 12 Asian, 8 Latin American, 9 European
Union (EU) countries, plus Bulgaria, Romania, Russia, South Africa and Turkey from 1980 to
2002.They conclude that the acceleration of growth immediately after liberalisation is found to be
often driven by an investment boom and a surge in portfolio and debt inflows. By contrast, the
quality of domestic institutions, the size of FDI inflows and the sequencing of the liberalisation
process are found to be important driving forces for growth in the medium to longer term.
26
Chinn and Itoh (2006) investigate whether financial openness leads to financial development over
the period 1980 to 2000 in 108 countries, including 30 Sub Saharan African (SSA) countries as
part of a broader set of developing or emerging countries. After controlling for legal institutions,
they find that a higher level of financial openness directly promotes the development of equity
markets and indirectly through its interaction with legal and institutional development; however,
the latter effect requires a certain threshold of institutional development. Their results are more
relevant to emerging economies than developing countries since they focused on equity markets.
Nonetheless, their study seems to support the argument that an argument legal and institutional
infrastructure is necessary for financial liberalization to be effective. They also find that trade
openness is a precondition for capital account liberalization and the development of the banking
system is required for equity market development.
Omoke (2010) in his own study concludes that in a period of financial liberalization, trade
openness and financial development have causal impact on economic growth. Others who find
similar results that financial development is important for economic growth are, Gallego and
Loayza (2002), Soukhakian (2005), and Okpara (2010).
O’Donnell (2001) and Chanda (2005) also consider the possibility of differing the effects of
capital account openness across countries. O’Donnell in this study examines the impact of capital
account openness using both IMF rule-based measures and quantitative-based measure of financial
openness. Using a standard set up, they find that the rule –based measure tends to be too coarse an
indicator of the degree of capital account liberalization, as it does not take into account the nature
of different types of controls. However, using the quantitative measure, he finds that capital
account openness does seem to speed up economic growth. However, like other researchers, he
finds that the benefits are not equal.
In another study, Klein and Olivei, (2001) analyse the impact of capital account liberalization on
growth and financial depth for a cross-section of countries over the period 1986-1995. They found
that countries with open capital accounts experienced a larger increase in financial depth than
countries with closed capital account, and through that channel, higher rates of economic growth
occurs. Also, Chinn and Ito (2005) do find a positive effect of financial openness on domestic
financial development if the institutional quality in the country is of a sufficiently high level.
27
On the other hand, some studies such as that of Eichengreen and Leblang, (2003) find a negative
relationship between financial openness and growth, while Grilli and Milesi-Ferretti, (1995), find
that financial openness does not affect growth. Using a cross section of countries, this study
considers average growth of per capita income for five non-overlapping five-year periods between
1966 and 1989. Their sample includes 61 countries, although, with 181 observations in one set of
regressions and 238 in another, not every country appears in each of the five sub periods. Their
results do not support the hypothesis that capital account liberalization promotes growth.
Following the financial crises in Asia, Russia and Latin America in the 1990s, some authors have
argued that capital account liberalization does not generate efficiency. Instead, liberalization
invites speculative hot money flows and increases the likelihood of financial crises with no
discernable positive effects on investments, output or any other real variable with non-trivial
welfare implications (Bhagwati, 1998; Stiglitz, 2002)
Rodrik (1998) questioned the effect of capital account liberalization on growth. In a sample that
includes almost 100 countries, developing as well as developed, he finds no significant effect of
capital account liberalization, as measured by Share, on the percentage change in real income per
capita over the period 1975 to 1989 in growth regressions that also include initial per capita
income, initial secondary-school enrollment rate, an index of the quality of governmental
institutions, and regional dummy variables. Likewise, he finds no relationship between capital
account liberalization and investment-to-income, or between capital account liberalization and
inflation.
Chanda (2001) suggests that the impact of capital account openness on economic growth may vary
with the level of ethnic and linguistic heterogeneity in the society, a proxy for the number of
interest groups. In particular he finds that capital controls lead to greater inefficiencies and lower
growth among countries with a high degree of ethnic and linguistic heterogeneity.
Edison, et al (2002) also finds little evidence of a relationship between capital account
liberalization and growth. Using a variety of econometric techniques and methodologies (i.e two
cross-sectional ones based on OLS and IV and one based on a dynamic panel data model using
GMM) and a new data set focusing on quantitative measures rather than rule-based measures, they
find that financial integration does not accelerate economic growth per se, even when controlling
28
for particular economic, financial, institutional, and policy characteristics. They do, however, find
that international financial integration is positively associated with real per capita GDP,
educational attainment, banking sector development, stock market development, the law-and-order
tradition of the country, and government integrity (low levels of government corruption).
Aizenman (2004) apply a two-step FGLS procedure for a panel of developing and OECD countries
for the years 1982-1998 using annual observations. He finds that de-facto financial openness
depends positively on lagged trade openness, and GDP/Capita. The budget surplus to GDP ratio
is occasionally significant and always negative for developing countries, but positive and
significant for the OECD countries. Including the corruption variable in his regressions also yields
negative and significant coefficients in almost all the iterations of the model he examined,
confirming Wei’s (2000) insight.
Klein and Olivei (2008) however, argue that the lack of a positive growth effect of financial
openness in developing countries is due to a missing effect of financial openness on financial
development for these countries.
Table 2.2: Summary of Some Related Empirical Works on Financial Openness and Growth
Study No. of
Countries
Liberalization
/Openness
Measure
Dependent Variable
and Estimation
Method
Results
Grilli &
Milesi-
Ferretti (1995)
61 Share Growth in income per
capita for five-year non-
overlapping
periods during 1971 –
1994 period. IV
estimation.
No Evidence of a significant effect
of Share on
growth of income per capita.
Quinn
(1997)
58 ΔQuinn,
between
1988 and
1958
Growth in income per
capita 1960 – 1989.Cross
Section, OLS.
ΔQuinn significantly raises
growth in income per
capita, though no regression
is presented with both
29
ΔCapital Controls and
ΔOpenness.
Kraay
(1998)
64, 94, or
117
Share;
Quinn;
or Volume
Growth in income
per capita over
1985 – 1997. Cross
Section.
OLS & IV. Samples
of 117 (Share); 94
(Volume); or 64
(Quinn).
No effect of Share or Quinn
on Growth. Coefficient
on Volume significant and
positive.
Rodrik
(1998)
About 100 Share Growth in income
per capita over
1975 – 1995. Cross
Section,
OLS.
No Evidence of a significant
effect of Share on
growth of income per capita
Klein &
Olivei
(2001)
67 Share Growth in income
per capita, 1976 –
1995. Cross
Section, IV.
Change in Financial
Depth (ΔFD ) as a
function of Share
and
then per capita
income growth as a
function of
instrumented
Significant effect of Share on
ΔFD, though results
Seem to be driven by OECD
countries in sample.
Significant effect of
instrumented values of ΔFD
And FD on growth.
30
Value of ΔFD (and
initial FD).
Arteta,
Eichengreen.
&
Wyplosz,
(2001)
51
to 59
Quinn in
Initial
Year; or
ΔQuinn
over
relevant
period
Growth in income
per capita 1973 –
81, 1982 – 87, 1988
– 92, or
Pooled for these 3
periods. Follows
Edwards (2001) but
with
OLS rather than
WLS and with
different
instruments.
Quinn significant for pooled
results but not for
Shorter subsamples. ΔQuinn
not significant.
Significant effect of
interaction of Quinn with
either
quality of law or openness
Bekaert,
Harvey &
Lundblad
(2001)
30
Emerging
markets
Official
Dates of
Stock
Market
Liberalizati
on
Growth rates in
income per capita
for various time
periods
between 1981 and
1997, resulting in
overlapping data.
Stock market liberalization
significantly contributes
to growth in income per
capita, with largest effects
shortly after liberalization
Chanda
(2001)
57
non-
OECD
Share Growth in income
per capita over
1975 – 1995. Share
interacted
with measure of
ethnic
heterogeneity.
Share significantly raises
growth in ethnically
heterogeneous countries and
significantly lowers it
in ethnically homogeneous
countries.
31
Edwards
(2001)
55 to 62
Quinn in
1988; or
ΔQuinn
1988 –
1973
Growth in income
per capita, 1980 –
1989. Cross
Section. WLS
(1985 GDP as
weight), IV. Also
uses interaction of
Quinn in
1988 and log (GDP
in 1980).
Quinn level significantly
raises GDP growth.
Interaction suggests that, at
low GDP, opening
Capital account may lower
GDP growth.
O’Donnell
(2001)
94 Share or
Volume
Growth in income
per capita over
1971 – 1994.
Regressions
include interaction
between FD and
Share, and Volume
and FD.
Neither Share nor interaction
of Share and FD
significant, but Volume
sometimes significant.
Edison,et al
(2002)
57 Share,CapSt
ocks, or
CapFlow
Growth in GDP per
capita over 1980–
2000,
Cross section (OLS
and IV); Panel
(GMM).
Regressions include
interaction
terms with financial
development, initial
Using a wide array of
measures, they find
no evidence that
international financial
integration accelerates
economic growth.
32
income, schooling,
institutional factors,
and macroeconomic
policies
Klein (2003) 84 or 52 Share or
Quinn
Growth in income
per capita 1976–96,
Cross section.
Specification
allows for
quadratic
interaction with
initial income.
Significant effect found for
middle-income
Countries but not for poor or
rich countries.
Hermes and
Lensink
(2003)
67
developing
Gross FDI
inflows to
GDP (1970-
1995)
Growth of real per
capita GDP
Positive significant
coefficient on interaction of
FDI with FD variables and
growth
Alfaro et al
(2004)
71 Net FDI
inflows to
GDP (1975-
1995)
Growth of real per
capita GDP
Robust to additional controls
and IV estimation and also
shows a positive significant
coefficient on interaction of
FDI with FD variables.
Bekaert et al
(2005)
95 De-jure
international
equity
market
liberalizatio
n (1980-
1997)
5-year average
growth rate of real
per capita GDP
Significant higher growth
gain post-liberalization for
countries with higher FD
33
Prasad et al
(2007)
83 Stock
liabilities,
gross and
net flow
liabilities to
GDP
Growth in real
sector value added
Countries with below median
FD showed negative
significant coefficient on
interaction of external finance
dependence of industry and
Financial Openness
Coricelli et
al (2008)
31 Euro-
Economies
and
Transition
Economies
Sum of
stock of
total foreign
assets and
liabilities as
a percentage
of GDP
Growth of real
GDP per capita
For financial development, the empirical analysis in Coricelli et al. (2008), using industry-level data for EU and transition countries, revealed that it indeed significantly contributed to growth and catching-up in transition economies. Similarly, financial integration seemed to play a comparably important role in the growth
performance of transition
economies.
Omoke
(2010)
Nigeria Direct
Credit,
Private
Credit and
Money
Supply
Financial Development Proxies
Domestic credit, Private
credit and broad money, as
percentages of GDP showed
no causal impact on growth.
NOTES: Share is proportion of years that IMF’s AREAR shows open capital accounts. Quinn is Quinn’s 0 – 4 measure of
capital account intensity.
ΔQuinn is change in value of Quinn 0 – 4 measures. Volume is measure of volume of capital flows. Cross Section refers to 1
observation per country. FD= Financial Development
Source: Author’s compilation based on Literature Reviewed and in line with Kose et al (2010) and Edison et al (2004)
34
2.3.2 Direction of Causality between Financial Openness and Economic Growth
Empirically, not many studies have been done on the actual direction of causality between financial
openness and economic growth. And for a developing economy like Nigeria, there seems to be a
dearth of such empirical studies. And where it exists, the results have been very inconclusive
(Osinubi and Amaghionyeodiwe 2010, Oyatoye, et al 2011 and Ogbonna,et al, 2012).
Our review on the direction of causality between financial openness and economic growth shall
therefore be based on disaggregated de facto financial openness and economic growth. This is
because different components of aggregate financial openness can be determined or caused by
different factors as argued by Ozdemir and Erbil (2008) and Aizenman and Noy (2009).
In this regard, Zhang (2001) looks at 11 countries on a country-by-country basis, dividing the
countries according to the time series properties of the data. Tests for long run causality based on
an error correction model, indicate a strong Granger-causal relationship between Openness to
capital flows (FDI) and GDP-growth. For six counties where there is no co integration relationship
between the log of FDI and growth, only one country exhibited Granger causality from Openness
to growth. Chowdhury and Mavrotas (2006) take a slightly different route by testing for Granger
causality using the Toda and Yamamoto (1995) specification, thereby overcoming possible pre-
testing problems in relation to test for co integration between series. Using data from 1969 to 2000,
they find that Financial Openness (proxied by FDI) does not Granger cause GDP in Chile, whereas
there is a bi-directional causality between GDP and Openness in Malaysia and Thailand.
De Mello (1999) looks at causation from Financial Openness (proxied by FDI) to growth in 32
countries of which 17 are non-OECD countries. First he focuses on the time series aspects of FDI
on growth, finding that the long run effect of FDI on growth is heterogeneous across countries.
Second, de Mello complements his time-series analysis by providing evidence from panel data
estimations. In the non-OECD sample he finds no causation from FDI to growth based on fixed
effects regressions with country specific intercepts, and a negative short run impact of FDI on GDP
using the mean group estimator.
Choe (2003) in another panel study, use method developed by Holtz-Eakin et al. (1988) in a data
set of 80 countries to test for causality between openness to capital flows and growth. His results
35
show a bi-directional causality between FDI flows and growth, but he finds the causal impact of
FDI on growth to be weak. In their own study, Basu et al. (2003) address the issue of the two-
way link between growth and FDI. They find a co-integrated relationship between FDI flows and
growth using a panel of 23 countries while allowing for country specific co- integrating vectors as
well as individual country and time fixed effects. They argue that trade openness is a crucial
determinant of the impact of FDI on growth, as they find that long run causality is unidirectional
from growth to FDI in relatively closed economies, while two-way causality between FDI and
growth is evident in open economies, both in the short and the long run.
Christopoulos and Tsionas (2004) address the high frequency factors influencing the finance
growth nexus by using panel cointegration analysis. They find that the long-run causality runs
from financial development to growth. Bekaert et al. (2001, 2005) show that financial liberalisation
spurs growth by improving resource allocation and increases the accumulation rate.
Quinn and Toyoda (2008) argue that in analyzing the causal relationship between capital account
liberalization and economic growth, differing time periods and collinearity among independent
variables result in measurement errors. In this line, they employ both time series and cross sectional
analysis and GMM analysis for a long time period of 1955-2004. They observe a positive
relationship between financial openness and economic growth independent of the country’s
economic development level.
Abu-Bader and Abu-Qarn (2008) examine the causal relationship between financial development
and economic growth in Egypt during the period 1960-2001 using a trivariate VAR framework.
The paper employs four different measures of financial development (ratio of money to GDP, ratio
of M2 minus currency to GDP, ratio of bank credit to the private sector to GDP, and the ratio of
credit issued to private sector to total domestic credit). The paper suggests that the causality is bi-
directional. Moreover, the paper shows that the impact of financial development on growth is
through both investment and efficiency.
Ahmed (2008) in another study, use financial openness as proxy for financial liberalization and
adopted the fully modified OLS (FMOLS) and Granger-causality test to the estimate long-run
financial development-growth and causality relationships. He finds that financial development
36
exerted a negative impact on economic growth when private credit was used as a proxy, while the
relationship was positive but insignificant when domestic saving was employed. However, the
financial liberation index exerted a positive and significant impact on economic growth. He
concludes that both in the short run and long run, the Granger-causality test does not find evidence
of any causality relationship between financial development and GDP per capita growth.
Aizenman and Noy (2009) study the causality and endogenous determination of financial openness
and trade openness. They construct a theoretical framework leading to two-way feedbacks between
financial and trade openness. Their results show that one standard deviation increase in
commercial openness is associated with a 9.5% increase in de facto financial openness (% of
GDP). Similarly, they find that an increase in de facto financial openness has powerful effects on
future trade openness. De jure restrictions on capital mobility have only a weak impact on de facto
financial openness, while de jure restrictions on the current account have a large adverse effect on
commercial openness. The authors further investigate the relative magnitudes of these directions
of causality using Geweke’s (1982) decomposition methodology. They conclude that in an era of
rapidly growing trade integration, countries cannot choose financial openness independently of
their degree of openness to trade. Dealing with greater exposure to turbulence by imposing
restrictions on financial flows is likely to be ineffectual.
Moudatsou and Kyrkilis (2009) study the causal-order between capital inflows (proxied by inward
FDI) and economic growth using a panel data set for two different Economic Associations that is
EU (European Union) and ASEAN (Association of South Eastern Asian Nations) over the period
1970-2003. They investigate three possible cases in their paper (1) Growth-driven FDI, is the case
when the growth of the host country attracts FDI (2) FDI-led growth , is the case when the FDI
improves the rate of growth of the host country and (3) the two way causal link between them.
From the heterogeneous panel analysis they find the following: Regarding the EU countries the
results support the hypothesis of GDP -FDI causality (growth driven FDI) in the panel. Regarding
the ASEAN there is evidence that there is a two ways causality between GDP per capita and FDI
in the cases of Indonesia and Thailand while in the cases of Singapore and the Philippines FDI is
host country GDP growth motivated.
37
Sridharan et al (2009) in a study of Brazil, Russia, India, China and South Africa (BRICS)
economies, report bidirectional causation between openness to capital flows and GDP for Brazil,
Russia and South Africa and unidirectional (FDI leads growth) for India and China. Similarly, in
a panel study of China, Japan, India, South Korea and Indonesia Agrawal and Khan (2011) analyse
the impact of openness on GDP Growth using data for 1993 to 2011. Their results show a
unidirectional impact where FDI promotes economic growth, and further provides an estimate that
one dollar of FDI adds about 7 dollars to the GDP of each of the five countries investigated.
However, the study by Geogantopoulos and Tsamis (2011) report a unidirectional link running
from GDP to FDI but found no bidirectional causation between FDI and GDP in Greece.
Imene and Schalck (2010) study the causality relationship between financial flows/ development
and economic growth in Tunisia. They focused on the link between finance and growth according
to the maturity of financial systems. Their empirical study aimed to determine the direction of
causality and impact of the development of the Tunisian financial system on economic growth
within B-VAR framework. They find a reciprocal relationships between the ratio of investment on
GDP and the loans granted to private and public sectors. They conclude that the Tunisian economy
knew a long period of financial repression before starting several phases of liberalization. Thus,
the economic role of government is highlighted, over the pre-reforms period as well as during the
recent time.
Hanh (2010) use the Pedroni co-integration technique and the GMM estimator to investigate the
possible causal connection between financial development, financial openness and trade openness
in twenty-nine Asian developing countries over the period 1994-2008. They find an evidence of
bidirectional causality between financial openness and trade openness as well as between financial
development and trade openness. Furthermore, when they interact the financial development and
financial openness terms, they find on one hand, that the openness indicator (Gross Private Capital)
is positively related to the ratio of credit to the private sector to GDP (PRIVO) indicator, but does
not influence the ratio of liquid liabilities to GDP (LLY) indicator. On the other hand, they find a
bidirectional causality between FDI and PRIVO indicators and a unidirectional causality running
from LLY to FDI. They therefore argue that trade openness is necessary for attracting financial
openness (foreign capital flows) promoting the development of financial system. In the same vein,
38
financial openness and financial development is seen as an important condition for trade openness
to take place in developing countries.
Aslanoğlu and Deniz (2012) analyze the causal relationship between stability in openness and
growth for emerging economies, namely as, China, Brazil, India, Indonesia, Korea, Mexico,
Russia, Turkey and South Africa while selecting a developed country, the UK, as a benchmark.
They test for causality between stability in financial openness and growth under different
frequencies and find that stability in openness has a positive impact on economic growth. They
conclude that the impetus on growth is conditional on economy’s ability attract foreign funds and
adopt exchange rate stability.
Zakari et al (2012) examines the causal role of openness on growth (as measured by foreign direct
investment (FDI) and GDP), making a comparison between selected countries of Africa and Asia.
They utilized data for 30 countries, 15 each from Africa and Asia for the period 1990 to 2009.
First, they analyzed the aggregate data and later disaggregated the data into Africa and Asia in
order to assess the regional impact of FDI on economic growth. Their empirical results reveal that
FDI has positive relationship with GDP growth for both Africa and Asia. However, they report
evidence of one-way causality for Africa and no such evidence for Asia. They conclude that FDI
promotes economic growth and recommend for more openness of the economies.
Abdelhafidh (2013) investigates the direction of causality between financial openness, finance and
growth in North African countries over the period 1970-2008. The study use Trivariate VAR
models to disentangle the direct and indirect impact of financial development on growth,
distinguishing between domestic saving and foreign inflows. They also disaggregated foreign
inflows into FDI, portfolio investment, grants, and loans. The result indicates that economic
growth Granger-causes domestic saving in some of the countries. However, in Tunisia and
Morocco particularly, grants Granger-cause growth and while growth Granger-causes loans. Also,
they find that in Egypt, FDI, grants, short-term loans, long-term loans, bank loans, bilateral loans
and multilateral loans, all Granger-cause growth with a reverse causality running form growth to
foreign inflows. Furthermore, they show that in Algeria, grants and multilateral foreign loans and
39
bonds Granger-cause growth. Thus, they conclude that these results reveal that policy initiatives
and directions should be tailored to each case given the peculiarity of each country that was
investigated.
Ayadi and Arbak (2013) using a sample of northern and southern Mediterranean countries for the
years 1985-2009 investigates the causal relationship between openness, financial sector
development and economic growth. Their empirical results reveal that bank deposits and credit to
the private sector are negatively associated with growth. They argue that this result confirms
deficiencies in credit allocation in the region and also suggests weak financial regulation and
supervision. On the stock market side, they find that stock market size and liquidity play a
significant role in growth, especially when accounting for the quality of an institution. They also
argue that investment, whether domestic or in the form of FDI (openness to capital inflows),
contributes significantly to economic growth while low inflation and stronger institutions are key
growth factors. They however conclude that initial GDP has a persistently and significantly
negative impact on growth, which implies that poorer countries are catching up with richer
countries in terms of economic growth.
2.3.3 Financial Openness and Output Volatility
There are plethora of literature on financial openness and its effect on macroeconomic volatility.
However, understanding the effect of financial openness on output volatility is very important for
relatively poor countries such as Nigeria which are exposed to exchange rate and terms of trade
shocks owing to their dependence on basic commodities such as oil. The empirical relationship
between financial openness and output volatility is undeniable, making volatility a fundamental
development concern. A number of studies have tried to explain the nature and causes of this
relationship but there is no consensus yet.
Backus, et al. (1992) argue that, if most shocks are country-specific and transitory, financial
opening should lower consumption volatility while raising investment volatility. However, the
empirical literature cannot provide statistically significant evidence on the relationship between
financial openness and macroeconomic volatility (Razin and Rose, 1994). Kaminsky and
40
Schmukler (2003) observe that, although equity markets stabilize in the long run (i.e. in five years
or longer) if financial liberalization persists, the amplitudes of booms and crashes substantially
increase immediately following financial liberalization.
Kose, et al (2003) in a detailed study , provide a comprehensive examination of changes in
macroeconomic volatility in a large group of industrial and developing economies over the period
of 1960 − 1999. They find that on average, the volatility of consumption growth relative to that of
income growth has increased for more financially integrated developing economies in the 1990s.
They also report a threshold effect, where the adverse effects increasing financial openness
diminish for more developed countries. In another study, Prasad, et al (2003) compare the volatility
experiences of a sample of 22 more financially integrated developing countries and 33 less
financially integrated developing countries. They find that over the 1990s, the most financially
integrated have experienced some increase in consumption volatility while the less financially
integrated group of developing countries and the industrialized countries both experienced
average declines in consumption volatility relative to the previous decade.
Buch, et al (2005), using a panel dataset for OECD countries, find that the implications of financial
openness for business cycle volatility depend on the nature of the shocks and the link between
macroeconomic policy, financial openness, and business cycle volatility varies over time. They
further argue that developing economies are more vulnerable to external shocks due to some
structure features, e.g., limited diversification of foreign trade, sudden reversal of capital flows,
the small country size. These factors hamper the unbiased empirical estimation of the relationship
between financial openness and macroeconomic volatility. This is in line with Kose (2002) which
shows in a dynamic small-open-economy model that terms of trade shocks can explain a sizeable
fraction of volatility.
Hagen and Zhang (2006) develop a model of a small open economy and show that financial
openness has non-monotonic relationship with macroeconomic volatility. After pooling the
empirical data of countries with different degrees of financial openness, they conclude that
domestic financial frictions may explain the lack of strong empirical evidences on the significant
linear relationship between financial openness and macroeconomic volatility. In another study,
41
Giovanni and Levchenko (2006) show that countries that are more open to trade tend to be more
volatile than others. They argue that this the outcome of counteracting forces. Two mechanisms
lead to a positive relationship: traded sectors are more volatile than nontraded ones, and trade leads
to specialization in fewer sectors. But traded sectors are less correlated with the rest of the economy
and so can act hedging activities.
Prati and Tressel (2006) find that foreign aid volatility increases trade balance volatility and
depresses exports through a Dutch-diseases mechanism. They argue that these effects could be
mitigated by actively managing the central bank’s net domestic assets. Levchenko and Mauro
(2007), concerned with the detrimental effects of sudden stops, reveal that countries with a more
diversified portfolio of foreign liabilities and a higher share of foreign direct investment tend to
fare better during capital-flow reversals.
Loayaz and Raddatz (2007) use a semi-structural vector autoregressions on a panel of countries to
study how financial openness, trade openness, factor-market flexibility, product-market flexibility
and domestic financial development influence the impact of terms of trace shocks on output. They
find that financial openness and labour market flexibility appear to reduce the impact of external
shocks. On the other hand, they find that trade openness increase the output consequences of terms-
of-trade shocks, especially when domestic markets are not well developed. These findings are
consistent with the finding of Bruner and Ventura (2006), who studied how trade integration can
lead to financial instability using a model of endogenously incomplete market. In the study, they
argued that trade integration can have different effects depending on domestic financial markets.
If those markets are thin, trade integration destroys risk-sharing and lowers welfare. If those
markets are deep, trade integration allows for better risk-sharing, thus raising welfare.
Furthermore, empirical analysis by Aguiar and Gopinath (2007) suggest that Emerging Market
Economies (EMEs) are vulnerable to sudden stops in capital inflows, and that these economies are
twice as volatile as that of industrial countries. Not surprisingly, if one looks at historical data, the
volatility of developing countries’ real GDP is at least about thirty percent higher than that of the
OECD countries. This aggregate volatility, in turn, has severe implications at the micro-level, and
particularly for the poor who are the least equipped to weather these aggregate shocks and are
42
therefore likely to face the brunt of its harmful impact. Another empirical study by Calderón and
Yeyati (2007) suggests that inequality increases with economic volatility—a doubling in aggregate
income volatility (measured as the standard deviation of per capita GDP) leads to a 2.7 percent
increase in the Gini coefficient, a 2.4 percent reduction in the income share of the poorest quintile,
and a 1.1 percent increase in the income share of the richest quintile.
Also, Kose, et al (2008), in another study, examined the risk-sharing implication of financial
integration by focusing on the cross-country correlations of output and consumption. They find
that there is no evidence that financial globalization fosters increased risk-sharing across all
countries, including the developing countries.
Ahmed and Suardy (2009), examine the effects of both financial and trade liberalization on real
output and consumption growth volatility in Sub-Saharan Africa. They find that trade liberalization
is associated with higher output and consumption volatility while financial liberalization is
observed to increase the efficacy of consumption smoothing and stabilize income and consumption
growth. They conclude that there is evidence that good institutions help reduce inflation levels and
volatility, which in turn promote lower growth volatility.
Udah (2010) argues that Macroeconomic uncertainty plays a key role in determining investment
behaviour in developing countries. Uncertainties arise from high and unstable inflation rate,
unstable fiscal deficits, overvaluation or depreciation and exchange rate misalignment.
Macroeconomic uncertainty or instability could also arise from political instability or poor
macroeconomic management. When the future is highly uncertain, investors take a ‘wait’ and ‘see’
attitude. At the microeconomic level firms may decide to limit their capacity in the face of
uncertainty in demand conditions, which leads to reduced investment capacity.
Mougani (2012) analyzes the impact of financial integration on economic activity and
macroeconomic volatility in Africa within the financial globalization contexts. The results of the
empirical analysis show that the impact of external capital flows on growth seems to depend
mainly on the initial conditions and policies implemented to stabilize foreign investment, increase
domestic investment, productivity and trade, develop the domestic financial system, expand trade
43
openness and other actions aimed at stimulating growth and reducing poverty. The analysis also
shows that financial instability was particularly severe as from the nineties. The instability was
more pronounced in the case of portfolio investments than in foreign direct investment because of
the longer-term relationship established by the latter.
Orji, et al (2013) in a recent study investigate the relative impacts of the uncertainty of
macroeconomic variables on investment using Generalized Autoregressive Conditional
Heteroscedasticity (GARCH) model. Their findings reveal the existence of long run relationship
between some of the macroeconomic variables and investment. And also, that the uncertainty of
most of the macroeconomic variables impact negatively on investment in Nigeria.
Conclusion
The arguments, simulations, and evidence in the foregoing papers seem not to agree on the exact
causal direction between financial openness and growth and also on the impact of financial
openness on growth and output volatility. Since the above issues that were raised in the literature
review still remains largely unsettled, this thesis has done more empirical investigations to unravel
the answers in the Nigerian context.
2.4 Shortcomings of Previous Studies
1. The issue of adopting an adequate measurement of financial openness remains unsettled in the
literature. Most of the papers used either de jure or de facto measures but in this study we are
adopting the two measures to see the financial openness measure that will be more robust and
significant for the Nigerian economy. This robustness check allows us to establish whether the
effect of openness on growth and volatility is equally strong for various measures of
liberalization.
2. The above studies did not investigate direction of causality between Financial Openness in
detail. And where efforts were made to do so, they were not based on the disaggregated de facto
financial openness measure in the Nigerian context as we have done in this study. (See for
44
example Ayadi and Arbak (2013), Choe, 2003, Osinubi and Amaghionyeodiwe 2010, Oyatoye,
et al 2011l and Ogbonna,et al, 2012).
3. Most of the papers did not address country specific issues. Most of the studies were based on
cross-country analysis (for example see Quinn and Toyoda, 2008; De Mello, 1999). But in this
study, we are using Nigeria as a case study. Thus, engaging in a detailed country-specific
analysis.
4. Previous authors that ran cross country regressions on the effects of financial openness on
growth and volatility got conflicting results because of the number of countries estimated,
measure of liberalization used, time frame and other factors.
5. Several related studies reviewed neglected the role of institutional factors in their regression
estimation and analysis. However, in this study we are incorporating the role of governance in
our model to account for the peculiar institutional cum socio/political environment upon which
this research is based.
45
CHAPTER 3
METHODOLOGY
3.1 Theoretical Framework
The endogenous growth theory was constructed from the shortcomings of the neoclassical model
of economic growth, with Arrow (1962), Romer (1986) and Lucas (1988) being the key
contributors. In neo-classical growth models, the long run rate of growth is exogenously
determined by either the savings rate (as in the Harrod-Domar model) or the rate of technical
progress (as in the Solow Model). However, the source of the savings rate and the rate of
technological progress cannot be explained (Ghatak and Siddiki, 1999). Endogenous growth
models attempt to explain a greater proportion of observed growth as well as why different
countries experience different growth rates. They generally use the neoclassical model but allow
the production function to exhibit increasing returns to scale, focus on externalities and assume
that technological change, although important, is not necessary to explain long-run growth.
In trying to resolve the contending issues surrounding the neo-classical model, the endogenous
growth theorists construct macro-economic models out of micro-economic foundations. Thus,
households are assumed to maximize their utilities, subject to some budget constraints, while firms
maximize profits. In this sense, the most important aspect is usually attributed to innovation (the
invention of new technologies) and the human capital. The engine for growth can be as simple as
a constant return to scale production function (the AK model) or more complicated set ups with
spill-over effects (spill-overs are positive externalities, benefits that are attributed to costs from
other firms). For instance, Pagano (1993) uses an endogenous growth model which incorporates
human capital (L) in his study of financial markets, liberalization and growth. This is because
financial liberalisation leads to increase in the quality of human capital by financing education to
less endowed households in the society as Gregorio (1996) explains.
The endogenous growth theory holds the view that human capital is one of the main sources of
economic growth and development. This is a very important argument in the developing countries
due to the abundance of labour. The model put forward by Pagano (1993) predicts that financial
46
liberalization and openness will lead to increase in: (a) saving and investment; (b) the proportion
of saving that goes to investment and (c) the efficiency of investment as a result of improvement
competition as well as availability of information regarding the investment projects.
Using an AK version of endogenous growth model Pagano (1993) postulates that the three factors
aforementioned in turn increase the rate of economic growth. The extended model predicts that
there is an additional efficiency gain caused by the accumulation of human capital as a result of
financial liberalisation. To explain the model, assume that aggregate output is a linear function of
aggregate capital stock.
Yt = AKt - - - - . -(3.1)
where Yt is aggregate output, Kt is the aggregate capital stock and t is time. This production
function represents a competitive economy with the presence of externality or spill-over effects
(Ghatak and Siddiki, 1999). Each firm faces constant returns to scale, but the economy as a whole
shows increasing returns to scale with respect to Kt.
Furthermore, suppose that the population is constant and the economy produces a single
commodity which can either be consumed or be invested. Also, assume that the rate of amortisation
of capital stock is zero and gross investment is:
It = Kt+1 - Kt
Kt+1 = It + Kt - - - - - - - (3.2)
This is assumed to be a closed economy with only one-sector and no government. If we assume
that financial intermediaries channel a proportion φ of saving, St, to investment, It (i.e. a proportion
(1 - φ) of saving is lost through the process of intermediation and does not go directly to
investments. On the basis of this, the capital / money market equilibrium condition can be
expressed as:
ψSt = It - - - - - - (3.3)
Using equations (3.1) and (3.2), the growth rate (g) at time t+1 can be written as:
47
gt+1 = (Yt+1 – Yt)/Yt = (AKt+1 – AKt)/AKt = Kt+1/Kt – 1
gt+1 = (It + Kt)/Kt -1 = It/Kt = AIt/AKt - - - - - (3.4)
where gt+1 is the growth rate of output at time t+1 and the steady state is defined as :
Kt = Kt+1 = K; Yt = Yt+1 = Y; gt = gt+1 = g. Substituting equation (3.3) into
equation (3.4) the steady state growth rate (g) can be written as follows:
g = A (I/Y) = Aψs - - - - - - (3.5)
where s is S/Y. Taking the logarithms of equation (3.5), it can be expressed as:
Ln g = Ln A + Ln ψ + Ln s - - - - - (3.6)
Equation (3.6) shows the growth rate as a linear function of its determinants and channels through
which financial liberalization policies affects growth (A, ψ, s.) Our empirical model is therefore
based on this relationship.
3.2 THE MODELS
3.2.1 MODELS FOR OBJECTIVE 1
1. Modeling the Impact of Financial Openness on Economic Growth
Here, following Ozdemir and Erbil (2008) we employ two different measures of financial
openness. The first category refers to the de facto measure of financial openness. This measure is
price-based. Following Aizenman (2004 and 2008) and Aizenman and Noy (2009), the de facto
measure of financial openness can be used as a variable to measure the actual observed outcomes
of the enforcement of existing regulations on financial flows.
The second category is the de jure measure of financial liberalization. De jure measures are quality
based measures which concentrate on events such as changing regulations and the response of the
monetary authorities to financial flows.
Growth Regression Model with De facto and De Jure Financial Openness Measures
Equation (3.6) distinguishes three channels: ψ, s and “A” (Improvement in financial
intermediation, Savings, efficiency of capital stock), through which financial liberalisation policies
48
could influence economic growth. Using endogenous growth theory, this study examines a
modified version of the growth model used by Ozdemir and Erbil (2008) where the growth rate of
real GDP per capita is regressed on other financial sector indicators and other variables. Others
who have used similar models include Fowowe (2002) and Owusu (2012). But we differ by
including the De jure (FODJ) variable using the Chinn-Ito Index and by adding institutional
(Governance) Index. Thus, accounting for the peculiar political/ institutional environment upon
which this research is based.
Expanding Equation (3.6) and adding other relevant variables of interest we have:
The De facto Financial Openness and Growth Equation
퐿푌푃퐶 = 훼 + 훼 퐿푃푆퐶 + 훼 푅퐼푁푇푅 + 훼 퐻푀퐿 + 훼 퐿푀퐾푇퐶퐴푃 + 훼 퐹푂퐷퐹 + 훼 푅퐸푋퐶푅
+ 훼 퐼푁푆푇 + 휇 … … … … … … … … … … … … … … … … … . … … … . (3.7푎)
The De jure Financial Openness and Growth Equation
퐿푌푃퐶 = 훼 + 훼 퐿푃푆퐶 + 훼 푅퐼푁푇푅 + 훼 퐻푀퐿 + 훼 퐿푀퐾푇퐶퐴푃 + 훼 퐹푂퐷퐽 + 훼 푅퐸푋퐶푅
+ 훼 퐼푁푆푇 + 휇 … … … … … … … … … … … … … … … … … . … … … . (3.7푏)
where:
훼 . … …훼 =Parameter estimates
L= Natural log operator
µ= Error term
The above growth equation also has the following variables:
LYPC= Real GDP per capita growth rate (proxy for economic growth) following Edison (2002),
Alfaro et al (2004), Bekaert et al (2005), Coricelli et al (2008), Odhiambo (2009) and Owusu
(2012).
PSC= Credit to the private sector
This captures the improvements in the banking sector. It is expected that improvements in financial
intermediation will affect economic growth positively (Levine 2008)
49
RINTR= Real Interest Rate
Liberalisation of interest rate, according to McKinnon-Shaw hypothesis, leads to increase in
savings then increase in investments and ultimately leading to increase in economic growth. Using
a simple aggregate production function framework, Montiel (1995) shows that interest rate
liberalisation can alter the economic growth rate through three main channels: (i) increase in
investment resulting from the increase in savings rate; (ii) improvement in the efficiency of capital
stock and (iii) improvement in the financial intermediation.
HML= Human Labour (Proxied by Secondary school enrolment rate)
To improve the efficiency of capital requires human effort and this has been captured by including
capital stock (K) and a labour factor (L) in equations (3.6 and 3.7a&b). This is because the
endogenous growth theory posits that human capital is one of the main sources of economic
growth, especially in the developing countries. Human Labour (HML) and especially trained
labour, is expected to enhance productivity by giving incentives for innovation (Owusu, 2012).
The measure for labour is proxied by the secondary enrolment rates, which is defined as the ratio
of the number of enrolment at secondary schools to the total population (Shabhaz et al., 2008).
MKTCAP= Market Capitalization
This represents the total market capitalization of All Shares traded on the floor of the Nigerian
Capital Market within the period under review. Capital Market Liberalization has been emphasized
in the literature as one of the core areas of financial liberalization. Thus, we expect a positive
relationship between capital market liberalization and growth. Beck et al. (2000) in their study
outline three key stock market indicators in measuring size, activity, and efficiency. The ratio of
stock market capitalization to GDP (MC) measures the size of the stock market because it
aggregates the value of all listed shares traded in the stock market. They emphasize that one can
assume that the size of the stock market is positively correlated with the ability to mobilise capital
and to diversify risk. To measure stock market liquidity/activity and efficiency, they also used the
value of stock traded to GDP variable (VT) and Turnover Ratio (TR) respectively.
FODJ= De jure Financial Openness measured by Chinn-Ito Index. We use this index because of
its wide acceptability and it is available for a long period (up to 1970-2011) for over 182 countries
50
of the world including Nigeria. As earlier stated, the construction of the Chinn-Ito index is based
on the first principle component of four binary variables in IMF’s Annual Report on Exchange
Arrangements and Exchange Restrictions (AREARER) and it takes higher values for more open
financial regimes. These four variables are defined as follows: K1 is the variable that indicates the
presence of multiple exchange rates; K2 is the variable that indicates restrictions on current account
transactions; K3 is the variable that indicates the restrictions on capital account transactions; and
K4 is the variable that indicates requirements of the surrender of export proceeds1.
FODF = Financial Openness de facto measures. Here we use total capital flow as a ratio of GDP
to capture our degree of de facto Financial Openness. The sum of FDI, portfolio investments and
other investments make up the capital flows, (Aizenman and Noy, 2009). According to the World
Bank, “Gross private capital flows are the sum of the absolute values of direct, portfolio, and other
investment inflows and outflows recorded in the balance of payments financial account, excluding
changes in the assets and liabilities of monetary authorities and general government”.
In line with the endogenous theory, we also expect a positive relationship since this variable also
captures capital stock/ effects of external investment inflows (Sanchez-Robles and Bengoa-Calvo,
2002).
REXCR=Real Exchange Rate. We expect a negative relationship with growth since a rise in
foreign currency against the local currency affects foreign exchange demand which equally affects
capital imports and exports, investments and growth ultimately. Ozdemir and Erbil (2008)
INST= Institutional Quality Index (Proxy for Governance)
This variable helps us to measure the socio-political environment in which this study is based. We
measured this Index based on the data collected by the World Bank and other relevant bodies like
Political Risk Group for different countries including Nigeria. We expect sound governance which
is exemplified by respect for the rule of law to contribute positively to economic growth.
1For more detailed discussion on the construction of the Chinn-Ito Index see Chinn and Ito (2012). We justify the use of this index owing to its wide acceptability and availability.
51
THE AUTO REGRESSIVE DISTRIBUTED LAG (ARDL) MODEL
The Auto Regressive Distributed Lag (ARDL) Model which uses a bounds test approach based on
unrestricted error correction model (UECM) was employed here to measure the impact of
Financial Openness on Economic Growth and to test for a long run relationship among the relevant
variables. This model was developed by Pesaran and Pesaran (1997) and used by Pesaran, et al
(2001); Masron (2009); Owusu (2012), among others. The main advantage of this approach lies in
the fact that it can be applied irrespective of whether the variables are I (0) or I (1). This approach
also allows for the model to take a sufficient number lags to capture the data generating process in
a general-to-specific modelling framework. Although, a dynamic error correction model (ECM)
can be derived from ARDL through a simple linear transformation, Banerjee et al., (1998) and
Pesaran et al., (2001), have introduced bound testing as an alternative to test for the existence of
cointegration among the variables. The bounds test procedure is merely based on an estimate of
unrestricted error correction model (UECM) using ordinary least squares estimator. Tang (2003)
argues that the UECM is a simple re-parameterization of a general ARDL model. Also following
Shrestha and Chowdhury (2007), to illustrate the ARDL modelling approach, the unrestricted error
correction model of equation (3.7a&b) respectively is:
∆퐿푌푃퐶 = 훼 + 훿 퐿푌푃퐶 + 훿 퐿푃푆퐶 + 훿 푅퐼푁푇푅 + 훿 퐻푀퐿 + 훿 퐿푀퐾푇퐶퐴푃 + 훿 퐹푂퐷퐹
+ 훿 퐸푋퐶푅 + 훿 퐼푁푆푇 + 훼 ∆퐿푌푃퐶 + 훽 ∆퐿푃푆퐶 + 훾 ∆푅퐼푁푇푅
+ 퓂 ∆퐻푀퐿 + 휉 Δ퐿푀퐾푇퐶퐴푃 + ℊ퓃 ∆퐿퐹푂퐷퐹 퓃 + 휙 Δ푅퐸푋퐶푅
+ Ω퓏 ∆퐼푁푆푇 퓏 + ὠ… . . (3.8푎)
52
∆퐿푌푃퐶 = 훼 + 훿 퐿푌푃퐶 + 훿 퐿푃푆퐶 + 훿 푅퐼푁푇푅 + 훿 퐻푀퐿 + 훿 퐿푀퐾푇퐶퐴푃 + 훿 퐹푂퐷퐽
+ 훿 퐸푋퐶푅 + 훿 퐼푁푆푇 + 훼 ∆퐿푌푃퐶 + 훽 ∆퐿푃푆퐶 + 훾 ∆푅퐼푁푇푅
+ 퓂 ∆퐻푀퐿 + 휉 Δ퐿푀퐾푇퐶퐴푃 + 휑퓂 ∆퐹푂퐷퐽 퓂 + 휙 Δ푅퐸푋퐶푅
+ Ω퓏 ∆퐼푁푆푇 퓏 +ὠ… . . (3.8푏)
The terms with the summation signs in equations (3.8a&b) represent the Error Correction Model
(ECM) dynamics and the coefficients 훿 are the long run multipliers corresponding to long run
relationship (Poon, 2010). 훼 푎푛푑ὠ represent the constant and the white noise respectively. Δ is
the first difference operator while 푝푎푛푑푞are the lag length for the UECM. We conduct an F-test
for a joint significance by using ordinary least square (OLS) technique. As stated earlier, the
ARDL-UECM process will indeed enable us test the existence of long run relationships for the
model above.
3.2.2 Objective 2- Modeling the Direction of Causality between Financial Openness and
Economic Growth
Aizenman and Noy (2009) study the causality and endogenous determination of financial openness
and trade openness. They construct a theoretical framework leading to two-way feedbacks between
financial openness and trade openness. But we differ from Aizenman (2004) and Aizenman and
Noy (2009) by using a disaggregated measure of de-facto financial openness. Aizeman (2004) uses
a composite measure of financial openness but we use a disaggregated de facto openness (Ratio of
FDI to GDP) because various components of financial openness might be determined by different
variables.
Testing for the Direction of Causality between Financial Openness and Economic Growth
The issue of causality relationship as proposed by Granger (1963) is useful in analyzing how an
economic time series can be used to forecast another. Thus, a variable 푋 is said to Granger-cause
another series푌 , if given the past of푌 , past values of 푋 can help forecast푌 . According to Gujarati
(2004) the Granger causality test assumes that the information relevant to the prediction of the
53
respective variables in a given model is contained solely in the time series data of these variables.
Generally, it is important to note that since the future cannot predict the past, if variable 푋
Granger- causes variable푌 , then changes in 푋 should precede changes in푌 .Therefore, in a
regression of 푌 on other variables (including its own past values), if we include past or lagged
values of 푋 and it significantly improves the prediction of푌 , then we can say that 푋 granger-
causes푌 . A similar definition applies if 푌 granger-causes푋 . Thus, the model for our second
objective involves the following pair of regressions:
푌푃퐶 = 훼 퐹푂퐷퐹 + 훽푗 푌푃퐶 + 휇 ...(3.90)
퐹푂퐷퐹 = 휆 퐹푂퐷퐹 + 훿푗 푌푃퐶 + 휇 ...(3.91)
where:
훼 ,훽푗, 휆 ,훿푗 = 퐶표푒푓푓푖푐푖푒푛푡푠
FODF = De facto Financial Openness for Nigeria. As noted above Aizenman (2004) uses a
composite measure of financial openness but we use a disaggregated de facto openness (Ratio of
FDI to GDP) because various components of financial openness might be determined by different
variables. See Aizenman (2004:28)
YPC= Per Capita Gross Domestic Product (this captures the level of economic growth in the
country). A country with high and sustained growth rate is more likely to attract more capital
inflows. Thus, a positive relationship is expected between growth rate and financial openness. A
stable economic growth rate signifies a good economic performance and therefore is more
attractive to foreign investors (Sahoo, 2006)
In equations (3.90) and (3.91) it is assumed that the disturbances 휇 and 휇 are uncorrelated. In
the case of the causality equations above, we can distinguish four cases:
54
1. Unidirectional Causality from FODF to YPC is indicated if the estimated coefficients on the
lagged FODF in (3.90) are statistically different from zero as a group (i.e.∑휇 ≠ 0) and the set
of estimated coefficients on the lagged YPC in (3.91) is not statistically different from zero(i.e
∑훿 = 0).
2. On the other hand, unidirectional causality from YPC to FODF exists if the set of lagged FODF
coefficients in (3.90) is not statistically different from zero (i.e ∑휇 = 0 ) and the set of the
lagged YPC coefficients in (3.91) is statistically different from zero (i.e ∑훿 ≠ 0).
3. Bilateral causality or Feedback is suggested when the sets of FODF and YPC coefficients are
statistically different from zero in both regressions.
4. Finally, independence is suggested when the sets of FODF and YPC coefficients are not
statistically significant in both regressions.
However, it is pertinent to note that the application of the standard Granger test requires that the
variables YPC and FODF be stationary. Further to this, we apply the Error Correction Model
(ECM) to examine the short-run and long-run equilibrium adjustment dynamics of the variables.
Thus, we specify our Granger causality model based on the ECM framework as follows:
∆푌푃퐶 = 훼 ∆퐹푂퐷퐹 + 훽푗 ∆푌푃퐶 + 휛 Ѱ + 휇 ...(3.92)
∆퐹푂퐷퐹 = 휆 ∆퐹푂퐷퐹 + 훿푗∆푌푃퐶 + 휛 휑 + 휇 ...(3.93)
Where ∆푌푃퐶 and ∆퐹푂퐷퐹 are established first-differenced stationary, co-integrated time series.
Ѱ푡−1 and 휑푡−1 are lagged values of the error term and must be stationary if the first-differenced
∆푌푃퐶 and ∆퐹푂퐷퐹 series are co-integrated. The inclusion of Ѱ푡−1 and 휑푡−1 differentiates the
error correction model from the usual Granger causality regression.
3.3 JUSTIFICATION OF THE MODELS
One of the reasons for adopting the Autoregressive Distributed Lag (ARDL) methodology is that
it has been favoured by many researchers in recent years owing to its wide applicability
irrespective of whether the underlying regressors are purely I(0), purely I(1) or mutually co-
55
integrated. Engle – Granger (1987) co-integration analysis and Johasen (1991) maximum
likelihood are the most commonly used co-integration methodologies but due to the high predictive
power of the ARDL –Bounds testing methodology, it has become the choice of many researchers
in co-integration analysis in recent times (Shrestha and Chowdhury, 2007)
Another reason for using the ARDL approach is that is has been found to be more robust and
performs better for finite samples than other co-integration techniques. Furthermore, according to
Banerjee et al (1993), a dynamic error correction model (ECM) can be established from ARDL
through a simple linear transformation. The ECM integrates the short run dynamics with the long
run equilibrium without losing long run information. Shrestha and Chowdhury (2007) have also
argued that using the ARDL-Bounds testing approach mitigates against problems resulting from
non-stationary time series. Narayan and Narayan (2005) have also shown that the ARDL –Bounds
testing methodology based on the unrestricted error correction model (UECM) has numerous
advantages over the traditional co-integration methods.
The Granger causality model which we specified to address our second objective has also been
widely used in economic literature. The application of the standard Granger test involving two
variables 푋푡and 푌푡 requires that the variables be stationary. However, following Gupta and Komen
(2008), since most economic variables are non-stationary in level forms, the standard Granger
causality test is conducted using regressions based on differenced stationary variables. This
differencing process may throw away some useful long-run information about causal relationships
among the variables. It is therefore advisable to apply the ECM framework to examine the direction
of causality among the variables (Gupta and Komen 2008).
This advanced framework as provided by Granger (1983, 1986) and Engle and Granger (1987,
1991) suggests a more elaborate test of causality which is applied within the co-integration and
error-correction model (ECM). Engle and Granger (1991) proved that if the variables are
integrated of order I (1) and co-integrated, the standard Granger causality test is not appropriate,
as it does not consider the error correction term, which corrects the disequilibrium in the short-run.
This framework allows for a causal linkage between two variables which emanates from a common
trend or long-run equilibrium relationship. As it were, such causality may not be detected by the
standard Granger test which analyses short-run information given by the past changes in a variable,
56
푋푡 which helps in explaining current changes in another variable 푌푡.Another usefulness of the ECM
framework lies in its ability to detect the possibility of reverse causality or bidirectional causality.
It is also important to note here that in the ECM framework, the possibility of finding no causality
in either direction (as obtainable in the standard granger test) is ruled out. This is because as long
as 푋푡 and 푌푡have common trend (co-integrated), causality must exist in one direction at least.
3.4 UNIT ROOT TESTS
It is important to check each time series variable for stationarity or unit root before conducting the
co-integration test on specified models. The unit root test has to be conducted first because without
it, if the regression analysis is conducted in the traditional way and time series variables are found
to be non-stationary, the result will be spurious. Here we use the Augmented Dickey Fuller (ADF)
for the unit root tests. The ADF is unit root test for time series. It is shown in the equation below:
∆푌 = 훽 + 훽 푡 + 훿푌 + 훼 ∆푌 + 휀 ... . .(3.94)
where 푌 is the variable in question, 휀 is white noise error term and ∆푌 = (푌 −푌 ),
∆푌 = (푌 −푌 ), etc.
These tests are used to determine whether the estimated δ is equal to zero or not. The number of
lagged difference terms to include is often determined empirically, the idea being to include
enough terms so that the error term in (3.94) is serially uncorrelated. Fuller (1976) has compiled
cumulative distribution of the ADF statistics by showing that if the value of the calculated ratio of
the coefficient is less than critical value from ADF statistics, then Y is said to be stationary. Note
that we shall also conduct this test for our GARCH models specified below in sub-section 3.6.2.
3.5 COINTEGRATION TEST AND ESTIMATION PROCEDURE
3.5.1 THE ARDL-UECM PROCEDURE
There are various estimation techniques that can be adopted to evaluate co-integration
relationships among macro-economic variables. For instance, Engle and Granger (1987) can be
used for univariate co-integration analysis as well as Philips and Hansen (1990) procedure while
57
Johansen (1988) and Johansen and Juselius (1990) technique can be used for multivariate co-
integration analysis. The full information for maximum likelihood co-integration approach has
also been compiled by Johansen (1995).
However, this study adopts the newly proposed autoregressive distributed lag (ARDL) approach
popularized by Pesaran and Shin (1995), Pesaran et al (1996) and Pesaran et al (2001). We use the
ARDL – Bounds testing methodology to estimate the specified models and empirically analyse
the long run relationship and the dynamic interactions between the relevant variables. First, a test
of stationarity will be conducted and completed, and then the ARDL – Bounds testing approach to
co-integration analysis follows.
The ARDL-Bounds testing approach based on unrestricted error correction model (UECM)
technique also involves two stages. The first stage is to estimate the ARDL model of interest by
ordinary least square (OLS) in order to test for the existence of a long run relationship among the
relevant variables. This is done by constructing an unrestricted error correction model (UECM)
and then testing whether the lagged levels of the variables in each of the equations are statistically
significant or not. In other words, whether the null hypothesis of no long term relationship is
rejected or accepted. To achieve this, a Wald test (f-statistics version for bound-testing
methodology) for the joint significance of the lagged levels of the variables, i.e. testing the null
hypothesis against the alternative is performed. If the F-statistics is above the upper critical value,
the null hypothesis of no long run can be rejected, irrespective of the orders of integration for the
time series.
Alternatively, if the statistics fall below the lower critical values, then the null hypothesis cannot
be rejected. However, if the F-statistics falls between the upper and lower critical values, then the
result is inconclusive. In this case the asymptotic distribution of the Wald test (F-statistics) is non-
standard under the null hypothesis of no co-integration between the variables of interest,
irrespective of whether the explanatory variables are purely I(0) or I(1).
The general UECM model is tested downwards sequentially by dropping the statistically non-
significant first differenced variables for each of the equations to arrive at a “goodness of fit”
equation using general-to-specific strategy. Once the long run relationship or co-integration has
58
been established, the second stage of testing involves the estimation of the long run coefficients
(which represents the optimum order of the variables after selection by AIC or SBC) and then
deriving the associated error correction in order to calculate the adjustment coefficients of the error
correction term (Masih, et al, 2008). Therefore, the short run effects are captured by the
coefficients of the first differenced variables in the UECM model. Having done this, there is also
the need to perform a series of diagnostic tests on the stochastic properties established model. This
is because the existence of a long run relationship does not necessarily imply that the estimated
coefficients are stable (Bahmani-Oskooee and Brooks, 1999). This therefore, involves testing of
the residuals (i.e homoscedasticity, non- serial correlation, etc), as well as stability tests (i.e
Ramsey RESET and Cumulative Sum of Squares (CUSUM) tests) to ensure that the estimated
model is statistically robust.
It is also pertinent to note here that one of the arguments against the ARDL is that the estimators
will be inefficient and biased (or even inconsistent) in the presence of autocorrelation of the
disturbances (Feeny, 2005). However, Pesaran and Shin (1999) show that appropriately modifying
the orders of the ARDL model is adequate to simultaneously correct for residual serial correlation
and the problem of endogenous regressors, thus giving ARDL an advantage over other approaches
to cointegration. This is also justified by Harris and Sollis (2003) and Constant and Yue (2010).
In addition, Pesaran and Shin (1999), argue that endogeneity problems are addressed in this
technique by modeling the ARDL with the appropriate lags, thus correcting for both serial
correlation and endogeneity problems. Jalil et al (2008) in their study also show that endogeneity
is less of a problem if the estimated ARDL model is free of serial correlation. In this approach,
Khan el al, (2005) equally argue that where all the variables are assumed to be endogenous, the
long run and short run parameters of the model can be estimated simultaneously.
3.5.2 THE GRANGER CAUSALITY-ECM PROCEDURE
The steps involved in implementing the Standard Granger causality test based on our second
objective are as follows:
59
1. We regress current YPC on all lagged YPC terms and other variables, if any, but we do not
include the lagged FODF variables in this regression. This is the restricted regression. From
this regression we obtain the restricted residual sum of squares, RSSR.
2. Now we run the regression including the lagged FODF terms. This is the unrestricted regression.
From this regression we obtain the unrestricted residual sum of squares, RSSUR.
3. The null hypothesis is H0:∑훼 = 0 that is, lagged FODF terms do not belong in the regression.
4. To test this hypothesis, we apply the F test given by Gujarati (2004) namely,
F = (RSSR − RSSUR)/m . . . . (3.95)
RSSUR/(n− k)
which follows the F distribution with m and (n− k) df. In the present case m is equal to the number
of lagged FODF terms and k is the number of parameters estimated in the unrestricted regression.
5. If the computed F value exceeds the critical F value at the chosen level of significance, we reject
the null hypothesis, in which case the lagged YPC terms belong in the regression. This is another
way of saying that FODF causes GDP.
6. We repeat Steps 1 to 5 to test model (3.93), that is, whether YPC causes FODF. In addition to
this we shall also estimate the Granger causality model based on the error correction mechanism
(ECM) as specified in models (3.92) and (3.93). This follows an important theorem, known as the
Granger representation theorem, which states that “if two variables Y and X are cointegrated, then
the relationship between the two can be expressed as ECM” (Gujarati 2004: 825)
3.6 THE GARCH MODEL FOR OBJECTIVE 3
3.6.1 Modeling the impact of Financial Openness on Output Volatility
Our methodological framework here draws from the seminal work of Engle (1982), in which he
introduced the Autoregressive Conditional Heteroscedasticity (ARCH) to capture the issues of
volatility in financial time series analysis. Bollerslev (1986) introduced the Generalised
Autoregressive Conditional Heteroskedasticity (GARCH) which is an extension of Engle’s
original work. For the analysis of the impact of financial openness on output volatility, we employ
the volatility of GDP growth rate as a measure of output volatility and also adopt the GARCH
Model. Among others, we explore the relationship between the two measures to capture the
response of output volatility to both de jure and de facto openness. Greater financial liberalization
60
and larger capital inflows are expected to increase output volatility (Neumann and Ron, 2008).
However, the relationship may be more complicated in that financial liberalization and capital
inflows may not move in tandem or step-for-step with one another. For example, a country may
have limited access to foreign capital even if is financially open. Conversely, a country that is not
deemed to be financially open to capital by de jure measures may, in fact, have large capital flows
due to the circumvention of these capital controls. Thus, we explore the impact of the de jure and
de facto measures within the Nigerian context.
Before specifying the GARCH model, it is necessary to begin briefly with a review of the ARCH
and GARCH models. Engle (1982) developed a new class of stochastic process; the
Autoregressive Conditional Heteroskedasticity (ARCH) model which is a process where the
conditional variance is a function of lagged squared residuals. As earlier stated also, Bollerslev
(1986) introduced the Generalised Autoregressive Conditional Heteroskedasticity (GARCH)
which is an extension of Engle’s original work. It allows the conditional variance to be a function
of the lagged variance; i.e. it allows for both autoregressive and moving average (ARMA)
components in the heteroskedasticity variance. He showed that the GARCH model allows a better
representation of the volatility process while being more parsimonious.
3.6.2 THE MODEL
Lensik (2002) argued that the principal directions in evaluating volatility or uncertainty are: (i)
standard deviations of the variables, (ii) dispersion of the unpredictable part of a stochastic process,
(iii) Generalized autoregressive conditional heteroscedasticity (GARCH) model of volatility. Also,
according to Ahmed (2009), and Popov (2011) output volatility can be calculated as the standard
deviation of change in real output or real GDP. However we follow Fang and Miller (2012), to
specify a GARCH (1, 1) model of output volatility. This is a multivariate modeling approach of
GARCH where other explanatory determinants of output volatility are included in the variance
equation with lags.
We therefore specify our fundamental GARCH volatility process as:
휎 = 훼 + 훼 휇 + 훼 휎 .......(3.96)
61
which says that the conditional variance of 휇at time t depends not only on the squared error term
in the previous time period [as in ARCH(1)] but also on its conditional variance in the previous
time period (Gujarati, 2004).
where 휎 equals the conditional variance (squared variance), given information available at time t.
훼 is the constant, and the 훼 refers to a first order ARCH term (i.e., news about volatility from the
previous period) and 훼 a first order GARCH term (i.e., persistent coefficient). The conditions that
훼 ≥ 0,훼 ≥ 0,푎푛푑훼 + 훼 < 1ensure the positive and stable conditional variances of 휇 . The
sum 훼 + 훼 , measures the persistence of shocks to the conditional variances. To estimate output
variability, we take the conditional standard deviation of RGDP growth rate (Y) in GARCH (1, 1)
order as specified in equation (3.97)
To estimate the determinants of output volatility, we include the lagged values of our growth rate
and our de jure/ de facto measures of openness in the variance equation, while controlling for other
exogenous shocks. Thus, we specify our new conditional variance equation as:
휎 푌 = 훼 + 훼 휇 + 훼 휎 + 훿 Ž ....(3.97)
where “Ž” is a vector of explanatory variables that could determine or influence output volatility.
These explanatory variables included in the variance equation are: (1) FODJV which is the de jure
financial openness volatility variable from Chinn-Ito (2012) (2) FODFV which is the de facto
financial openness volatility variable measured from gross capital flows. (3) TOTV which stands
for Terms of trade volatility. This is used as a proxy for external risk premium. (4) EXRV is
Exchange Rate Volatility. (5) INFV which is inflation volatility. According to Aizenman (2008)
and Ahmed (2009) the values of these volatility variables can also be calculated from their standard
deviations.
3.6.3 MODEL JUSTIFICATION
The generalized autoregressive conditional heteroscedastic (GARCH) model has gained a lot of
attention in the literature since the introduction by Bollerslev (1986). Various works on testing for
or modeling of time-varying volatility of stock market returns and other economic variables have
been dominated by this model. The model includes past variances in the explanation of future
62
variances, which allows to capture the patterns usually exhibited by many financial time series
such as volatility clustering and large kurtosis.
GARCH model allows for both autoregressive and moving average components in the
Heteroskedastic variance. Conditional volatility measures such as GARCH is a better proxy for
measuring volatility than a measure of unconditional volatility like moving average standard
deviations or sample variability. GARCH models make use of future as well as past information
on its construction. Serven (2003) pointed out that sample variability does not amount to
uncertainty, except when events are unpredictable. He stressed that sample variability may
overstate uncertainty by including not only truly unpredictable innovations to the variables of
interest, but possibly (cyclical) movement partly predictable from their own past. Greene, (2005)
further stressed that another important usefulness of the GARCH specification is that it allows the
variance to evolve over time in a way that is much more general than the simple specification of
the ARCH model. Byrne and Davis (2002) also pointed out that conditional volatility measures
such as the ARCH or GARCH model highlights periods of concentrated volatility or volatility
clustering which might be expected to maximize uncertainty. Hsieh (1989) found that GARCH (1,
1) model worked well to capture most of the stochastic dependencies in the times series. Based on
tests of the standardized squared residuals, he found that the simple GARCH (1, 1) model did
better at describing data than previous models he used.
3.6.4 METHOD OF ESTIMATION
The GARCH (1, 1) is a generalization of the ARCH (q) model proposed by Engle (1982) as a way
to explain why large residuals tend to clump together, by regressing squared residual series on its
lag(s). However, empirical evidence shows that high ARCH order has to be selected in order to
catch the dynamics of the conditional variance. Bollerslev (1986) proposed the Generalized ARCH
(GARCH) model as a solution to the problem with the high ARCH orders. The GARCH reduces
the number of estimated parameters from an infinite number to just a few. According to Brook and
Burke (2003), the lag order (1, 1) is sufficient to capture all the volatility clustering that is present
in a data ceteris paribus.
The GARCH (1, 1) estimation process involves some steps. The first step is to examine the time
series properties of the data before applying other appropriate modeling procedures. Secondly we
63
need to estimate the model to obtain the residuals from the regression with which to test for ARCH
and GARCH features. The third step involves regressing the squared residual series and
conditional variance on their lags and on the other explanatory variables in the model.
3.7 DATA SOURCES
The data used in the study covers annual time series data from 1986 to 2011. We used Eviews
interpolation technique to convert them to quarterly series2. The sources of the data include various
issues of the Central Bank of Nigeria (CBN) statistical bulletins/ financial reports and National
Bureau of Statistics (NBS) publications. Others include: World Bank Publications and data from
organizations like Political Risk Services (PRS) Group, etc.
3.8 ECONOMETRIC SOFTWARE
Models one and two were estimated with Stata 11&13, while model three was estimated with
Eviews 6.0 econometric software package.The suitability of Eviews and Stata is enhanced by their
interactive nature, which makes them user-friendly, and time efficient in terms of output and
robustness of statistics generated. They are suitable for different kinds of regression model
estimations and data analysis.
22 As obtainable in time series econometrics, we converted our data to quarterly series in order to have sufficient number of observations to run our models. Cointegration analysis requires large number of observations which our annual data may not provide sufficiently due to the number of years studied.
64
CHAPTER FOUR
PRESENTATION AND ANALYSIS OF RESULTS
In this section, we present the empirical results and analysis based on the specified models. As
discussed earlier, before we go ahead with the ARDL bounds testing, we shall first of all test for
the stationarity of all the variables that are going to be used in the analysis to ensure their order of
integration. That is, whether they are of order I (0) or I (1) stationary.
4.1 UNIT ROOT TESTS
Unit root tests and the order of integration
Tables 4.1 presents the summary of the unit root test results for the series in levels and in first
differences. The ADF lag length was selected automatically by Akaike Information Criteria
(AIC).The result indicate that apart from Log(PSC), HML and REXCR which is integrated of
order zero, all other variables were non-stationary since their absolute value of ADF statistic
exceeded the critical value only at first difference. Furthermore, the results in Table 4.1 indicate
that most of the variables become stationary at first difference and this enabled the use of the error
correction model in the autoregressive framework.
Table 4.1 Summary of ADF Unit root test results of the series
Variable Mackinnon Critical Values
Level ADF Test Stat
1st Difference ADF Test Stat
Order of Integration
Log(YPC) -2.601 1.723 -3.286* I(1) YPC -3.600 -3.514 4.595* I(1) Log (FODF) -3.452 -3.220 -6.973* I(1) FODF -3.222 -2.084 4.223* I(1) FODJ -3.451 -2.352 -3.603* I(1) Log(PSC) -3.286 -3.505* I(0) Log(MKTCAP) -2.601 2.087 -3.946* I(1) HML -3.150 -3.465* I(0) RINTR -3.452 -3.192 -5.123* I(1) REXCR -2.891 -3.604* I(0) INST -2.600 0.544 -3.728* I(1)
NOTE: * indicates significant at 5%, probability levels Source: Computed by the Author
65
The results of the stationarity tests show that most of the variables are non-stationary at level.
These results are shown in Table 4.1 above. Having established the vector of variables of concern,
the order of integration and stationarity of all the series was conducted using the Augmented
Dickey-Fuller (ADF) principal of establishing unit root. The ADF test was conducted on variables
in order to determine their stationary nature and those found non stationary were differenced to get
rid of the stochastic trend, a phenomenon associated with time series data.
4.2 Bounds test
To select the appropriate lag length for the first differenced variables, we adopted a general-to-
specific approach using an Unrestricted Vector Autoregressive means of Schwarz Bayesian
Criterion (SBC). The results however, show a maximum of 2 lag lengths. As argued by Pesaran
and Pesaran (1997), variables ‘in first difference are of no direct interest’ to the bounds
cointegration test. Hence, any result that supports cointegration in at least one lag structure
provides evidence for the existence of a long-run relationships. The calculated F-statistic together
with the critical bounds values are also reported. The ARDL bounds test is based on the assumption
that the variables are I(0) or I(1) as shown above in the unit root table.
We chose a maximum lag order of 2 for the conditional ARDL vector error correction model by
using the Akaike Information Criteria (AIC). The calculated F-statistics are reported in Table 4.2
when each variable is considered as a dependent variable (normalized) in the ARDL regressions.
From these results, it is clear that there is a long run relationship amongst the variables when Log
(YPC) is the dependent variable because its F-statistic (4.60) is higher than the upper-bound critical
value (3.50) at the 5% level. This implies that the null hypothesis of no cointegration among the
variables is rejected.
Table 4.2 Bound test for the estimation with De facto Financial Openness Variable
Dependent Variable F- Statistics Decision Log(YPC) 4.60 Co-integration Log(FODF) 5.25 Co-integration Log(PSC) 4.97 Co-integration Log(MKTCAP) 6.85 Co-integration HML 3.06 No Co-integration RINTR 5.94 Co-integration REXCR 3.90 Co-integration INST 1.55 No Co-integration
66
Table 4.3 Bound test for the estimation with De Jure Financial Openness Variable
Dependent Variable F- Statistics Decision Log(YPC) 4.55 Co-integration FODJ 3.99 Co-integration Log(PSC) 2.09 No co-integration Log(MKTCAP) 4.38 Co-integration HML 3.93 Co-integrated RINTR 8.63 Co-integration REXCR 3.91 Co-integration INST 1.29 No co-integration
Critical values;
The value of our F-statistic is 4.50 and 4.55, and we have (k + 1) = 8 variables (YPC, FODF/FODJ,
PSC, MKTCAP, HML, RINTR, REXCR, and INST) in our model. So, when we go to the Bounds
Test tables of critical values, we have k = 7.
Critical Values
Table CI (iii) on p.300 of Pesaran et al. (2001) is the relevant table for us to use here. We haven't
constrained the intercept of our model, and there is no linear trend term included in the ECM. The
lower and upper bounds for the F-test statistic at the 5% significance level is [2.32, 3.50], i.e. I (0)
= 2.32 and I (1) = 3.50. As the value of our F-statistic exceeds the upper bound at the 5%
significance level, we can conclude that there is evidence of a long-run relationship between the
two time-series (at this level of significance or greater).
4.3 Estimation Results
Table 4.4: The ARDL Model for the De facto Financial Openness (Long-run) Dependent Variable Log (YPC) Variables Coefficient Std. Error T-statistics Probability Constant -0.46292 0.1874 -2.47 0.016 Log(YPC(-1)) -0.11487 0.0256 -4.47 0.000 Log(FODF(-1)) .030031 0.0129 2.32 0.023 Log(PSC(-1)) -0.04457 0.0196 -2.27 0.026 Log(MKTCAP(-1)) 0.08475 0.0220 3.85 0.000 HML(-1) 0.00241 0.0024 0.98 0.331 RINTR(-1) 0.00081 0.00055 1.45 0.152 REXCR(-1) -0.00077 0.0002 -3.75 0.000 INST(-1) 0.00836 0.0099 0.84 0.402
67
Short-run D(Log(YPC(-1)) 0.25674 0.1035 2.48 0.015 D(Log(YPC(-2)) 0.15798 0.0907 1.74 0.086 D(Log(FODF)) 0.022213 0.0217 1.02 0.312 D(Log(PSC)) -0.052936 0.1010 -5.24 0.000 D(Log(MKTCAP)) 0.08320 0.0363 2.29 0.025 D(HML) 0.02664 0.0096 2.76 0.007 D(RINTR(-1)) -0.00058 0.00078 -0.74 0.459 D(REXCR) -0.00092 0.00047 -1.93 0.057 D(INST(-1)) -0.04186 0.0377 -1.11 0.271 R-squared = 0.6453
Adj R-Squared = 0.5674 F-Statistics = 8.29 F-prob = 0.0000
Results of diagnostic tests X2 Statistics Probability Breusch-Godfrey LM test for autocorrelation
0.607 0.4357
White Heteroskedasticity 2.96 0.0851 Ramsey RESET Test
3.32 0.0739
The ARDL results above depicts the following process; Long-run (1, 1, 1, 1, 1, 1, 1, 1) and Short-
run (1, 2, 0, 0, 0, 0, 1, 0, 1). However, it is important to note that the long run elasticities or
coefficients can then be generated from the ARDL-UECM by using the estimated coefficients of
the one lagged independent variables, multiplied by a negative sign, and divided by the estimated
coefficient of the one lagged dependent variable (Bardsen, 1989 and Tang, 2003). The short run
coefficients are then derived from the estimated coefficient of the first differenced variable in
ARDL-UECM models (Poon, 2010). We applied these methods in calculating the long run and
short impact for the de facto and de jure estimated results. However, the ECM results showing the
short run dynamics for the parsimonious ARDL models are presented in tables 4.5 and 4.7
respectively.
Thus, from table 4.4, we see that the long-run multiplier between Log (YPC) and Log (FODF) is
- (0.030031/ -0.11487) = 0.26. In the long run, an increase of 1 percent in Log (FODF) will lead
to an increase of 0.26 percent in Log (YPC). In addition, the long-run multiplier between Log
68
(YPC) and Log (PSC) is -(-0.04457/ -0.11487) = -0.38, implying that in the long run, an increase
of 1 percent in Log(PSC) will lead to a decrease of 0.38 percent in Log (YPC). The long-run
multiplier between Log (YPC) and Log (MKTCAP) is - (0.08475/ -0.11487) = 0.73. This means
that in the long run, an increase of 1 percent in Log (MKTCAP) will lead to an increase of 0.73
percent in Log (YPC). The long-run multiplier between Log (YPC) and RINTR is - (0.00081/ -
0.11487) = 0.007. This means that in the long run, an increase of 1 percent in RINTR will lead to
an increase of 0.007 percent in Log (YPC).And the long-run multiplier between Log (YPC) and
REXCR is - (-0.00077/ -0.11487) = -0.006.Thus, in the long run, an increase of 1 unit in REXCR
will lead to a decrease of 0.006 percent in Log (YPC). Also the long-run multiplier between Log
(YPC) and INST is - (0.00836/ -0.11487) = 0.07
The short run and long run results reported in Table 4.4 clearly show that the de facto financial
openness (FODF) has a positive short run and long run impact on the economic growth (YPC) in
Nigeria. The coefficient of de facto financial openness is positive, as expected, as well as
statistically insignificant and significant in the short run and long run respectively. This suggests
that 1% increase in de facto financial openness leads to an increase of 0.02% in economic growth
in the short run and 0.26% in economic growth in the long run. This supports previous studies such
as Fratzscher and Bussierre (2004), Coricelli et al (2008), Loyayza and Ranciere (2006) to mention
a few which found long run relationship between economic growth and de facto financial
openness. However, those who found contrasting results include; Rodrick (1998), Eichengreen and
Leblang, (2003), Klein and Olivei (2008) among others.
Other variables included in the model such as, Market Capitalization (MKTCAP) and Human
Labour (HML), are also statistically significant and positively related to Economic growth in
Nigeria. Real interest rate is also found to have a positive relationship with economic growth.
Although this is not very significant but the results support the McKinnon-Shaw hypothesis, i.e.
in the long run interest rate liberalisation will ultimately lead to rapid economic growth.
It is also observed that the coefficient of credit to the private sector (PSC) has a negative sign both
in the short run and long run. This is contrary to expectation. However, this corroborates Obamuyi
(2009) which finds a negative relationship between private sector credit and economic growth.
The study attributes this finding to the fact that private sector credits are mainly used by some
69
borrowers to buy and sell instead of investing it into productive activities. Again, it has also been
discovered that many bank managers simply issue loans to their cronies and family members who
use the funds for other purposes rather than investing them productively. The coefficient may
suggest that 1% increase in the volume of credit to the private sector leads to a reduction of 0.05%
and 0.38% in economic growth in the short run and long run respectively.
Our institutional quality variable which represents governance and rule of law also shows some
interesting results. In the short run it reveals a negative relationship with economic growth but in
the long run we see a positive relationship between the two variables. Thus, in the short run 1%
change in the quality of institution will lead to 0.04% reduction in economic growth. While in the
long run, 1% change in the quality of institutions will affect economic growth positively by 0.07%.
This result attests to the fact that the present style of governance among the leaders has serious
negative impact on the growth of the Nigeria. This is exemplified by the fact that the principle of
rule of law is not respected, corruption has been enthroned in several leadership quarters, and there
is no internal democracy even among the political parties. When the quality of governance and
institution is weak, it simply translates to corruption, embezzlement of state funds meant for
infrastructural development, and several other anti -socio/economic outcomes. This finding
supports the study of Gupta, et al (2001), Tanzi and Davoodi (1997) among others.
In order to get the parsimonious model, we estimated the model by OLS, constructed the
residuals series, and then fitted a regular (restricted) ECM:
Table 4.5 Parsimonious ARDL-ECM for de facto financial openness
Dependent Variable D(Log(YPC)) Variables Coefficient Std. Error T-statistics Probability Constant 0.01716 0.0072 2.37 0.020 D(Log(YPC(-1)) 1.03055 0.1470 7.01 0.000 D(Log(YPC(-2)) -0.17376 0.0992 -1.75 0.084 D(Log(FODF)) 0.03096 0.0188 1.64 0.105 D(Log(PSC)) -0.36393 0.0801 -4.54 0.000 D(Log(MKTCAP)) 0.10294 0.0311 3.30 0.001 D(HML) 0.02424 0.0074 3.27 0.002 ECMt-1 -0.90778 0.1766 -5.14 0.000
70
R-Squared = 0.5617 Adj. R-squared = 0.5287 F-Statistics = 17.03 F-prob. = 0.0000
Results of diagnostic tests
X2 Statistics Probability Breusch-Godfrey LM test for autocorrelation
2.690 0.1010
White Heteroskedasticity 6.97 0.8083 Ramsey RESET Test
3.89 0.8083
From table 4.5, we notice that the coefficient of the error-correction term (ECMt-1) is negative and
very significant. This is what we would expect if there is co-integration and long run relationship
between economic growth (log (YPC)) and other regressors. The magnitude of this coefficient
implies that nearly 91% of any disequilibrium between log (YPC) and other variables is corrected
within one period (one quarter). The ECM results also show that a change in de facto financial
openness (FODF) is associated with a positive change in economic growth (Log (RGDP)). Also,
the coefficient of D (Log (MKTCAP)) shows that a change in the stock market capitalization is
positively associated with change in economic growth and it is statistically significant at 5% level.
Furthermore, the coefficient of the change in the Human Labour (D (HML)) is positive and
statistically significant at 5% level. However, the coefficient of D (Log (PSC)) is negative and
statistically significant. This coefficient may suggest that the bulk of the credit extended to the
private sector by the banks and other financial institutions goes into mostly buying and selling of
imported finished consumer goods rather than production for domestic consumption in the real
economy and export to the outside world.
Table 4.6: The ARDL model for the De jure Financial Openness (Long-run)
Dependent Variable Log(YPC) Variables Coefficient Std.Error T-statistics Probability Constant .0148587 0.0873 0.17 0.869 Log(YPC(-1)) -0.10509 0.0250 -4.19 0.000 FODJ(-1) 0.02805 0.0117 2.38 0.019
71
Log(PSC(-1)) -0.03035 0.0199 -1.53 0.131 Log(MKTCAP(-1)) 0.08263 0.0216 3.82 0.000 HML(-1) 0.00069 0.0026 0.27 0.790 RINTR(-1) 0.00006 0.0004 0.15 0.884 REXCR(-1) -0.00049 0.0002 -2.31 0.024 INST(-1) 0.00357 0.0096 0.37 0.712
Short-run D(Log(YPC(-1)) 0.23126 0.0985 2.35 0.021 D(Log(YPC(-2)) 0.12609 0.0869 1.45 0.150 D(FODJ) 0.02641 0.0489 0.54 0.591 D(Log(PSC)) -0.056525 0.0971 -5.82 0.000 D(Log(MKTCAP)) 0.07409 0.0366 2.02 0.047 D(HML) 0.03196 0.0097 3.27 0.002 D(RINTR(-1)) -0.00029 0.0007 -0.39 0.701 D(REXCR) -0.00049 0.0005 -0.98 0.328 D(INST(-1)) -0.08324 0.0300 -2.77 0.007 R-squared = 0.6348
Adj R-Squared = 0.5600 F-Statistics = 8.49 F-prob. = 0.0000
Results of diagnostic tests X2 Statistics Probability Breusch-Godfrey LM test for autocorrelation
2.044 0.1528
White Heteroskedasticity 3.81 0.0509 Ramsey RESET Test 2.93 0.3084
The ARDL-UECM results above depicts the following process; Long-run (1, 1, 1, 1, 1, 1, 1, 1)
and Short-run (1, 2, 0, 0, 0, 1, 0, 0). Again, following our initial calculations for our de facto results
in table 4.4, we can also see from table 4.6, that the long-run multiplier between Log (YPC) and
(FODJ) is - (0.02805/ -0.1051) = 0.26. In the long run, an increase of 1unit in FODJ will lead to
an increase of 0.26 percent in Log (YPC). In addition, the long-run multiplier between Log(YPC)
and Log(PSC) is -( -0.03035/-0.1051) = -0.28, implying that in the long run, an increase of 1
percent in Log(PSC) will lead to a decrease of 0.28 percent in Log (YPC). Furthermore, the long-
run multiplier between Log (YPC) and Log (MKTCAP) is - (0.08263/ -0.1051) = 0.78. This means
that in the long run, an increase of 1 percent in Log (MKTCAP) will lead to an increase of 0.78
percent in Log (YPC). And in the long-run multiplier between Log (YPC) and REXCR is - (-
72
0.0005/ -0.1051) = 0.004. In the long run, an increase of 1 unit in REXCR will lead to a decrease
of 0.004 percent in Log (YPC).
The short run and long run results reported in Table 4.6 equally show that the de jure financial
openness (FODJ) has a positive short run and long run impact on the economic growth (YPC) in
Nigeria. The coefficient of de jure financial openness is positive, as expected, as well as
statistically insignificant and significant in the short run and long run respectively. The result
equally suggests that 1% increase in de jure financial openness leads to an increase of 0.02% in
economic growth in the short run and 0.26% in economic growth in the long run. This supports
previous studies such as Quinn (1997), Bekaert et al (2005), Chinn and Ito (2005 and 2006) to
mention a few which found long run relationship between economic growth and de jure financial
openness. However, Ozdemir and Erbil (2008) found a negative impact of de jure financial
openness measure on growth.
Other variables included in the model such as, Market Capitalization (MKTCAP) and Human
Labour (HML), also have positive relation with Economic growth in Nigeria and statistically
significant in Nigeria.
However, it is also observed here that the coefficient of credit to the private sector (PSC) has a
negative sign both in the short run and long run. This is contrary to expectation. But this has
confirmed that high interest rate and excessive government borrowing are making private credit
inefficient and detrimental to growth; and that public expenditure is crowding out private sector
investment. This also reveals the problem of huge non-performing loans, and corporate
governance deficiencies of some lending banks, supporting the finding of Abubakar and Gani
(2013) and Nkoro and Uko (2013).
In order to get the parsimonious model, we estimated the model by OLS, constructed the
residuals series, and then fitted a regular (restricted) ECM:
Table 4.7: Parsimonious ARDL-ECM for de jure financial openness Dependent Variable D(Log(YPC)) Variables Coefficient Std. Error T-statistics Probability Constant 0.01924 0.0073 2.61 0.011 D(Log(YPC(-1)) 1.20877 0.2028 5.96 0.000 D(Log(YPC(-2)) -0.28793 0.1165 -2.47 0.015
73
D(FODJ) 0.05619 0.0445 1.26 0.209 D(Log(PSC)) -0.40684 0.0797 -5.10 0.000 D(Log(MKTCAP)) 0.07527 0.0333 2.26 0.026 D(HML) 0.02057 0.0074 2.75 0.007 D(RINTR(-1)) 0.00028 0.0007 0.37 0.712 D(REXCR) -0.00025 0.0003 -0.68 0.496 D(INST) 0.06375 0.0292 2.18 0.032 ECMt-1 -0.8905 0.2109 -4.22 0.000 R-Squared = 0.5814 Adj. R-squared = 0.5349
F-Statistics = 12.50 F-prob. = 0.0000
From table 4.7, we also notice that the coefficient of the error-correction term (ECMt-1) is negative
and very significant. This is what we should expect if there is co-integration between log (YPC)
and other regressors. The magnitude of this coefficient implies that nearly 90% of any
disequilibrium between log (YPC) and other variables is corrected within one period (one quarter).
The ECM results also show that a change in de jure financial openness (FODJ) is associated with
a positive change in economic growth (Log (RGDP)). Also, the coefficient of D (Log (MKTCAP))
shows that a change in the stock market capitalization is positively associated with change in
economic growth and it is statistically significant at 5% level. Furthermore, the coefficient of the
change in the Human Labour (D (HML)) is positive and statistically significant at 5% level.
However, the coefficient of D (Log (PSC)) is negative and statistically significant. This coefficient
may suggest that the bulk of the credit extended to the private sector by the banks and other
financial institutions goes into mostly buying and selling of imported finished consumer goods
rather than production for domestic consumption in the real economy and export to the outside
world. Some of the credits also end up in the pockets of the cronies of the bank managers who
neither use the loans for productive purposes nor service the loans as at when due. Thus, we have
the perennial problem of bad loans that have become detrimental to the banking system’s ability
to purposefully finance private sector investments.
Table 4.8 Results of diagnostic tests
X2 Statistics Probability Breusch-Godfrey LM test for autocorrelation
2.501 0.1137
White Heteroskedasticity 4.24 0.0395 Ramsey RESET Test 5.58 0.1015
74
4.4 Further Interpretation and Explanations of Model Parameters
4.4.1 The long run model for economic growth and de facto financial openness (ARDL-
UECM)
Using a log model, the effects of financial openness on economic growth was modeled and results
were presented in Table 4.4. Observations made from the table indicate that from the long run part
of the ARDL-UECM, private sector credit and exchange rate affect economic growth negatively
while de facto financial openness indicator (FODF), market capitalization, interest rate,
institutional quality and human labour impacted positively on economic growth.
Further analysis of results of Table 4.4 indicate that, the positive association of financial openness
and GDP with market capitalization imply that size of the stock market dominated by money
lending organisations like banks, finance companies, insurance companies, micro-finance
institutions and other money suppliers boost economic growth. The capital base of these financial
institutions is also a key determinant of financial openness. This also implies that the financial
openness policies put in place as directed by the IMF in 1980’s have been favourable in increasing
the level of economic activities at the stock market, hence leading to an increase in market
capitalization.
Also from the model, interest rate is found to have a positive relationship with economic growth,
but this effect was found to be statistically insignificant in the long run. This result is expected
since interest rate is supposed to have a positive relationship with savings which is a key driver of
economic growth. This finding however, agrees partly with Perera’s case of financial liberalisation
where he found that interest rate and real gross domestic product impacted positively on money
demand while financial liberalization had negatively impacted on both M1 and M2 in a study of
impact of financial liberalization on money demand and economic growth in Sri lank (Perera,
2005).
The results above further indicate that financial openness positively affects economic growth while
private sector credit negatively affects economic growth. This agrees with the findings of
Odhiambo (2009) in a similar study carried out on South Africa.
75
4.4.2 The long run model for economic growth and de jure financial openness (the ARDL- UECM)
Also using a log model, the effects of de jure financial sector openness on economic growth was
modeled and results were presented in Table 4.6. Results from the long-run part of the model
indicate that de jure financial openness indicator (FODJ), market capitalization, real interest rate
and human labour impacted positively on economic growth, while private sector credit, real
exchange rate and institutional qualities affect economic growth negatively.
The results here clearly show that de jure financial openness, market capitalization and real interest
rate affect economic growth positively while private sector credit negatively affects economic
growth. The implications of these result outcomes have been discussed above. The model also
shows that about 64 percent variation in real gross domestic product is explained by the covariates
here considered. This is significant as indicated by its F-statistic of 8.49 and its probability of 0.00.
4.5 Interpretation of the Error Correction Models and Results for economic growth and de facto
financial openness
4.5.1 The short run dynamics and de facto financial openness
The short run dynamic model was estimated by the restricted ARDL ECM procedure. The levels
of the UECM ARDL model was estimated by OLS where the residuals series was constructed, we
then fitted a regular (restricted) ECM. The maximum lag was established by the minimum AIC
which minimizes the standard errors. The estimated OLS error correction terms measured the
transitory deviations from the steady state equilibrium value of each variable present in the long
run relationship. The coefficient of the error correction term in this case measures the speed of
adjustment from the short run to the long run equilibrium.
Parsimonious Restricted ARDL-ECM (1, 2, 0, 0, 0, 0, 0) Dependent Variable D(Log(YPC)) Variables Coefficient Std.Error T-statistics Probability Constant 0.01716 0.0072 2.37 0.020 D(Log(YPC(-1)) 1.03055 0.1470 7.01 0.000 D(Log(YPC(-2)) -0.17376 0.0992 -1.75 0.084 D(Log(FODF)) 0.03096 0.0188 1.64 0.105 D(Log(PSC)) -0.36393 0.0801 -4.54 0.000 D(Log(MKTCAP)) 0.10294 0.0311 3.30 0.001 D(HML) 0.02424 0.0074 3.27 0.002 ECMt-1 -0.90778 0.1766 -5.14 0.000
76
R-Squared = 0.5617 Adj. R-squared = 0.5287 F-Statistics = 17.03 F-prob. = 0.0000
Results from the table above suggest that the current value of financial openness has a positive
impact on economic growth (although it is not significant).But the first lag of GDP per capita,
current value of private sector credit, market capitalization and human labour significantly affect
economic growth, although the second lag of GDP per capita is insignificant. All variables are here
considered significant at 5 percent level. The coefficient of ECMt-1 (-0.908) is significantly
different from zero and bears the right sign thus validating the existence of cointegration in the
system. Thus, it indicates that when an external shock disturbs the equilibrium condition of
economic growth, about 91 percent of it is absorbed within one period (i.e one quarter in this
study).
In view of the table above and as regards significance of the model, the F-statistic and its
probability justify that it is highly significant and thus reliable. The model explains about 56
percent of the overall variations in the dependent variable.
Moreover, as earlier stated, the current values of financial openness were found here to be rightly
signed. This result implies that the effect of financial openness on economic growth is positive.
Real interest rate was found to have had a positive impact on economic growth in the long run
though this diverts from the short run effect. The empirical evidence from the long run analysis
are therefore in line with the findings of the Shrestha and Chowdhury (2007), Ghatak (1997)
Odhiambo (2009b) and Odhiambo (2009c) which found the positive effects of interest rate
liberalisation and argue that interest rates liberalisation leads to more savings, which ultimately
leads to increase in investment and economic growth.
The Procession to Parsimonious Model
It is important to note that the table and result above represent that of the parsimonious model. The
reduction process eliminated most of the insignificant variables without losing valuable
information. The regression results show that the goodness of fit in both models is satisfactory.
77
The F-statistics with its probability values of 0.000 indicate that, overall, the models are
significant. These results imply the rejection of the null hypotheses that all the right hand side
variables except the constant terms have zero parameter coefficients. The Breusch-Godfrey LM
test for autocorrelation of 0.10 does not point to any serious autocorrelation problems.
4.5.2 The short run dynamics of Economic Growth and de jure financial openness
The short run dynamics of economic growth show how the effects in the long run function of
economic growth adjusts period after period. The coefficient of the error correction term shows
the magnitude of this adjustment as presented in the table below.
Parsimonious Restricted ARDL-ECM (2, 1, 0, 0, 0, 1, 0, 0) Dependent Variable D(Log(YPC)) Variables Coefficient Std. Error T-statistics Probability Constant 0.01924 0.0073 2.61 0.011 D(Log(YPC(-1)) 1.20877 0.2028 5.96 0.000 D(Log(YPC(-2)) -0.28793 0.1165 -2.47 0.015 D(FODJ(-1)) 0.05619 0.0445 1.26 0.209 D(Log(PSC)) -0.40684 0.0797 -5.10 0.000 D(Log(MKTCAP)) 0.07527 0.0333 2.26 0.026 D(HML) 0.02057 0.0074 2.75 0.007 D(RINTR(-1)) 0.00028 0.0007 0.37 0.712 D(REXCR) -0.00025 0.0003 -0.68 0.496 D(INST) 0.06375 0.0292 2.18 0.032 ECMt-1 -0.8905 0.2109 -4.22 0.000 R-Squared = 0.5814 Adj. R-squared
= 0.5349 F-Statistics = 12.50 F-prob. = 0.0000
Source: Computed by the Author Analysis made with reference to the above table indicate that past values of real GDP affect current
values up to the second lag with significant values, while a one period lag of financial openness
affects current values of real GDP positively (though insignificant). Other covariates such as
private sector credit, market capitalization, human labour and institutional qualities affect
economic growth with significant values and varying magnitudes as can be seen from the table.
Of great importance is the coefficient of the error correction term here marked ECMt-1. As seen
from above two cases it bears the correct sign and it shows a very high adjustment towards
attainment of equilibrium condition. It validates the fact that cointegration exists between the
78
variables in the model and more so that if there is an exogenous effect that disturbs the equilibrium
level of the economy, about 90% is attuned in the first period.
The model explains about 58 percent variations in the model and it is highly significant as indicated
by the F-statistic. The Breusch-Godfrey LM test for autocorrelation indicates no evidence of serial
correlation in the residuals.
Further analysis made from the table indicates that financial openness is positive both in the short
run and long run. In the long run financial openness is significantly related to economic growth
but is insignificant in the short run. These findings are in line with the findings of Bwire (2007)
that financial openness proxied by financial development (FD) had significant positive effect on
economic growth since the liberalisation of the financial sector, using data from Bank of Uganda.
Moreover, the findings are in line with the findings of Mckinnon’s (1973) model and the financial
deepening approach by Edward Shaw (1973), where financial liberalisation acts as a catalyst to
growth through investment in high yielding projects resulting in an increase in real income. With
both de facto and de jure financial openness variables being positive and significant in the long-
run based on their ARDL- UECM results, the results indicate that financial openness is a beneficial
policy. This could be attributed to its ability to increase the financial base of the economy and
increase productive capital inflows.
4.6 ARDL- UECM and Short-run ARDL-ECM model diagnostic tests
Here, the emphasis is on testing the presence or absence of serial correlation in the residuals
generated from the models, Ramsey model specification test, heteroskedasticity test and stability
test.
4.6.1 Tests for serial correlation of residuals
The serial correlation tests of the residuals were based on the Breusch-Godfrey LM test for
autocorrelation. All the estimated models have their second order tests below them. Results from
the second order tests indicate no evidence of serial correlation in all the models.
4.6.2 Ramsey reset test
All the estimated models indicate no evidence of omitted variable problem in all the results. Thus
they passed the model specification test.
79
4.6.3 White Heteroskedasticity test
Also, most of our estimated models passed the white heteroskedasticity test in all the results.
4.6.4 Stability Tests
The stability of the long-run coefficient is tested by the short-run dynamics. Once the ECM models
of de jure and de facto financial openness equation have been estimated, the cumulative sum of
recursive residuals (CUSUM) tests are applied to assess the parameter stability (Pesaran and
Pesaran ,1997). The Graph below depicts the results for CUSUM tests. The results indicate the
absence of any instability of the coefficients because the plots of the CUSUM statistic fall inside
the critical bands of the 5% confidence interval of parameter stability. .
Figure 5. Cumulative Sum of Recursive Residuals (CUSUM) Test
4.7 GRANGER CAUSALITY MODEL RESULTS FOR OBJECTIVE 2
4.7.1 THE GRANGER CAUSALITY TEST
The Two variables were tested for their stationarity and found to be stationary at first difference.
The result of the granger causality test is presented below excluding the ECM values since our
main variables of interest here are financial openness (FODF) and economic growth (YPC). The
causality test results suggest a-bidirectional causation between Economic Growth (YPC) and the
de facto financial openness (FODF). Thus, we could represent this relationship as (FODF→YPC)
and (YPC→FODF). The probability of the F statistics is significant at 5 percent using a two-tailed
-30
-20
-10
0
10
20
30
90 92 94 96 98 00 02 04 06 08 10
CUSUM 5% Significance
80
test. This is a clear indication of the relative positive impact of financial openness on the economic
growth of Nigeria.
Table 4.9 Null Hypothesis Df F-Statistics Probability FODF does not Granger Cause YPC
2 11.384 0.003
YPC does not Granger Cause FODF
2 6.4938 0.039
This finding of bidirectional relationship between financial openness and economic growth
supports the results of Hanh (2010) and Chowdhury & Mavrotas (2006). The result of this
estimation (bidirectional causality) is also very informative in predicting how economic growth
will be affected if policymakers are to change financial openness policies and vice versa. Even
when we controlled for other macro variables, the model results showed no significant difference.
Thus, this evidence of bidirectional causality between financial openness and economic growth
suggests that financial openness is necessary for enhancing economic growth. In turn, economic
growth seems to be an important condition for financial openness to take place and also thrive in
Nigeria.
4.8 GARCH MODEL RESULTS FOR OBJECTIVE 3: Impact of Financial Openness on Output Volatility
4.8.1 THE GARCH MODEL
Unit root test
The result of the ADF test of stationarity is presented on Table 4.10 below. Here, Real exchange
rate was found to be stationary at level; while others, GDP growth rate (Y), terms of trade, de facto
financial openness, de jure financial openness and inflation were found to be stationary at first
difference at the 5% level for the ADF test.
81
Table 4.10a Summary of ADF Unit root test results of the GARCH series
Variable Mackinnon Critical Values
Level ADF Test Stat
1st Difference ADF Test Stat
Order of Integration
Y -3.600 -3.514 4.595* I(1) FODFV -3.222 -2.084 4.223* I(1) TOTV -2.5868 2.097 3.827* I(1) FODJV -3.451 -2.352 -3.603* I(1) REXRV -2.891 -3.604* I(0) INFV -3.455 -3.427 -4.286* I(1)
NOTE: * indicates significant at 5%, probability levels Source: Computed by the Author Table 4.10b Descriptive characteristics of the variables
Variables/ Statistics
Y FODFV TOTV REXRV INFV FODJV
Mean 2.199808 27.17365 104.4112 108.127 22.12626 2.030385 Median 2.3 27.07500 87.935 100.21 13.53 1.7 Maximum 8.89 36.73000 221.26 439.18 76.43 3.02 Minimum -4.21 16.35000 42.4 53.97 1.86 1.35 Std. Dev. 3.084427 4.728219 51.77508 50.61129 20.10329 0.599218 Skewness 0.098315 -0.047166 0.906833 4.036898 1.246872 0.582155 Kurtosis 2.370927 2.678203 2.467907 2.39166 3.144525 1.482444 Jarque-Bera 1.882386 0.487292 15.48087 2178.325 27.03845 15.85391 Probability 0.390162 0.783765 0.000435 0.00000 0.000001 0.000361
Computed by the author
As can be seen in Table 4.10b, four variables in the series are non-normally distributed. The null
hypothesis of normal distribution is rejected for TOTV, REXRV, FODJ and INFV at the 1% level,
and at the 5% level for the rest of the series. The mean and median of TOTV and REXRV are
positive and high above 100% respectively. This suggests that terms of trade and real exchange
rate especially at the beginning of each fiscal quarter were significantly positive and perhaps imply
that higher average values attract larger economic growth especially terms of trade.
The variables from the table also do not show evidence of fat tails, since the Kurtosis did not
exceed 3, which is the normal value. But there is little evidence of negative skewness, for de facto
financial openness volatility. These imply left and right fat tails, respectively. We can, therefore,
employ the GARCH model since there is no much kurtosis problem.
82
Confirmation of data as AR (1)
Before the estimation of the GARCH model the data was confirmed to be suitable for AR (1). The
correlogram test for the series is shows a slow decay in AC and a spike in the PAC on first
observation and then drops to zero/under zero indicates the AR(1). The first order autocorrelation
is 0.914, and they gradually decline to .0030 after 15 lags. These autocorrelations are not large,
but they are very significant. Though some are positive and others negative, which is expected in
most economic time series and yet is an implication of the GARCH (1, 1) model.
Then the above is followed by an estimation of Y, AR (1) using OLS in order to test the hypothesis
whether there is autocorrelation (and hence no ARCH) or not.
Table 4.11
Dependent Variable: Y Method: ML – ARCH Coefficien
t Std. Error z-Statistic Prob.
C -11.42835 14.27232 -0.800736 0.4233 AR(1) 0.981731 0.014924 65.78130 0.0000 Variance Equation C 0.317251 0.058825 5.393102 0.0000 ARCH(1) 1.650808 0.317355 5.201769 0.0000
R-squared 0.830084 Mean dependent var 2.181845 Adjusted R-squared 0.824935 S.D. dependent var 3.094039 S.E. of regression 1.294568 Akaike info criterion 2.867000 Sum squared resid 165.9148 Schwarz criterion 2.969320 Log likelihood -143.6505 F-statistic 161.2139 Durbin-Watson stat 1.871836 Prob(F-statistic) 0.000000 Inverted AR Roots .98
Arch test In order to test the ARCH effect we performed the ARCH-LM Heteroskedasticity test. Often a
“Ljung box test” with 15 lagged autocorrelations is used.
83
Table 4.12 =================================================================== F-statistics 4.74130 Probability 0.000003 Obs *R-squared 43.72930 Probability 0.000121 ========================================================================
The test p-values shown in the last column are all zero, resoundingly rejecting the null hypothesis
which says that “there is no ARCH”. On top of the p-values, out of the result of the
Heteroskedasticity test we see that, the ARCH-LM statistic is 4.74130 significant at the 5% level,
thus we conclude that there is ARCH effects.
Table 4.13 The GARCH Model Results
Mean Equation Coefficient Z-Statistics Probability Constant (µ) 2.373219 3.388535 0.0007 AR(1) 0.737731 11.38609 0.0000 Variance Equation Constant(ώ) 5.607304 12.26144 0.0000 ARCH (1) (α) 0.225949 0.558670 0.5764 GARCH (1) (β) 0.424017 1.612945 0.1068 Y (-1) (λ) -0.166743 -1.427170 0.1535 FODFV (-1) (δ) -0.033537 -0.762094 0.4460 TOTV (-1) (ƞ) -0.010199 -1.270723 0.2038 FODJ (-1) (ζ) -0.0406922 -0.702016 0.4827 REXR (-1) (ν) -0.007939 -0.758922 0.4479 INFV (-1) (ρ) -0.022769 -1.096954 0.2727 R-squared Adjusted R-squared
0.808468 0.787649
F-statistics 38.83371 Prob(F-statistics) 0.00000
Inverted AR Roots .74 Computed by the author
Analysis of the GARCH Model
From the result in the table above, the ARCH and GARCH coefficients (0.2259 and 0.4240) are
not statistically significant. The sum of these coefficients is 0.650 which indicates that shocks to
volatility have a persistent effect on the conditional variance. These shocks will have a permanent
84
effect if the sum of the ARCH and GARCH coefficients equals unity (that is, the conditional
variance does not converge on a constant unconditional variance in the long run).
Interpretation of the output volatility determinants
The coefficients of lag of GDP growth rate, de facto and de jure financial openness volatility, terms
of trade, real exchange rate, inflation in the GARCH (1, 1) measure the predictive power of
previous values of the variables on economic growth in Nigerian economy. As can be seen from
Table 4.13 the coefficients are all negative implying that a change in either of the variables in the
previous period in Nigeria reduces conditional output volatility this quarter. However, the findings
infer that financial openness and other macroeconomic variables in the model have not made any
significant impact on output volatility in Nigeria.
This finding is in line with Ramey and Ramey (1995) who in a detailed study documented an
empirical relationship that show that growth and volatility are negatively correlated. This is an
important result since it suggests that policies and exogenous shocks that affect growth may not
necessarily influence volatility.
Furthermore, the findings from the above GARCH result agree with existing studies such as; Razin
and Rose (1994) ; Easterly, Islam, and Stiglitz (2001) ; Buch, Dopke, and Pierdzioch (2002),
O’Donnell (2001) and Kose et al (2008), that were unable to document a clear positive empirical
link between openness and macroeconomic volatility. For example, Razin and Rose (1994) study
the impact of trade and financial openness on the volatility of output, consumption, and investment
for a sample of 138 countries over the period 1950–88. They find no significant empirical link
between openness and macroeconomic volatility. Easterly, Islam, and Stiglitz (2001) explore the
sources of macroeconomic volatility using data for a sample of 74 countries over the period 1960–
97. They find that a higher level of development of the domestic financial sector is associated with
lower volatility.
On the other hand, an increase in the degree of trade openness leads to an increase in the volatility
of output, especially in developing countries. Their results indicate that neither financial openness
nor the volatility of capital flows has a significant impact on macroeconomic volatility.
85
Buch, Dopke, and Pierdzioch (2002) use data for 25 OECD countries to examine the link between
financial openness and business cycle volatility. They report that there is no consistent empirical
relationship between financial openness and the volatility of output.
O’Donnell (2001) examines the effect of financial integration on the volatility of output growth
over the period 1971–94 using data for 93 countries. He finds that a higher degree of financial
integration is associated with lower (higher) output volatility in OECD (non-OECD) countries. His
results also suggest that countries with more developed financial sectors are able to reduce output
volatility through financial integration.
Arch test
In order to test the ARCH effect of the GARCH estimation result above, we performed the ARCH-
LM Heteroskedasticity test. Often a “Ljung box test” with 15 lagged autocorrelations is used.
Table 4.14 ===================================================================== F-statistics 2.22306 Probability 0.012845 Obs *R-squared 27.85531 Probability 0.022492 =====================================================================
However, the GARCH model did not assume a symmetric response of volatility to past shocks.
The test p-values shown in the table above are close to zero, rejecting the null hypothesis which
says that “there is no ARCH” at 5 % level. Also, the Q statistics on all the lags in the specification
do reject the null thus they support the hypothesis that the standardized residuals not serially
correlated.
However, we have tested further even longer lag (36 default lag) for the square residual and we
have observed that the Q-stat from lag 1 to lag 7 did not reject the null. But it rejected the null
hypothesis from lags 8 to 16 and failed again to reject at lag 17 Choosing shorter lag might result
to one failing to capture the lag order however, the longer lag one chooses the lower the power
will be.
Note: If there is no serial correlation, the autocorrelations and partial autocorrelations at all lags
should be nearly zero, and all Q-statistics should be insignificant with large p-values.
86
4.9 EVALUATION OF RESEARCH HYPOTHESES
Hypothesis 1(Ho1): Financial Openness has no significant impact on Economic Growth in Nigeria.
Decision Rule: Reject Ho1 if tcal>ttab, accept H0 otherwise.
Conclusion: From the result discussed in section 4.3 (tables 4.4 and 4.6) both de facto and de jure
financial openness have been found to have long run significant positive impact on
economic growth in Nigeria. Therefore, we reject the null hypothesis that Financial
Openness has no significant impact on Economic Growth in Nigeria.
Hypothesis 2 (Ho2): There is no significant direction of causality between financial openness and
economic growth in Nigeria.
Decision Rule: Reject Ho1 if tcal>ttab, accept H0 otherwise.
Conclusion: Following the result shown in table 4.9, we find that there a significant bidirectional
relationship between financial openness and economic growth in Nigeria. Thus, we
reject the null hypothesis (H0) and conclude that the there is a clear significant
direction of causality between financial openness and economic growth in Nigeria.
Hypothesis 3 (Ho3): Financial Openness has made significant impact on output volatility in
Nigeria.
Decision Rule: Reject Ho1 if tcal>ttab, accept H0 otherwise.
Conclusion: From the GARCH model results in table 4.14, we have seen that both measures of
financial openness have not made any significant positive contribution to output
volatility in Nigeria. Therefore, we reject the null hypothesis (H0) and conclude that
Financial Openness has not made significant impact on output volatility in Nigeria.
87
CHAPTER FIVE
SUMMARY OF THE FINDINGS, CONCLUSIONS AND RECOMENDATIONS
5.1 Summary of the findings
The study was set out to investigate the impact of financial openness on macroeconomic
performance in Nigeria. It established that the policies of financial openness and financial
liberalization introduced in Nigeria from late 1980s till present have made positive impact on
economic growth. However, our results suggest greater impact in the long run than in the short
run. As it were, one may deduce from the estimated models that de facto and de jure measurements
of financial openness have similar relationship with economic growth in the short run and long run
respectively.
Again, we also see from the results that financial openness in Nigeria have impacted positively
more on economic growth through market capitalization and negatively through private sector
credit, with the two being significant both in the short run and long run. Thus, there is significant
evidence that financial openness generally made positive contributions to economic growth within
the period under review. Nonetheless, an unexpected finding from the result is the negative sign
of private sector growth in both de jure and de facto financial openness growth results. This
supports the finding of Ayadi et al (2013) which indicates that credit to the private sector and bank
deposits are negatively associated with growth. This confirms deficiencies in credit allocation in
some countries and also suggests weak financial regulation and supervision. This result calls for
serious caution on the side of domestic financial sector managers to ensure that credit is targeted
at those sectors that are growth-enhancing. Generally, our finding agrees to the fact that capital
account openness, domestic financial market liberalization and stock market liberalization have
made different degrees of impact on the Nigerian Economy.
The results also support the findings of Ghatak (1997) in the Sri Lankan context as he explored
and found in a similar study, a positive impact of Financial Liberalization on economic growth of
Sri Lanka for the duration of 1950 to 1987. Furthermore, the findings of this study partly agree
with those of Odhiambo (2009) that financial development which results from interest rate reforms
did not cause investment and economic growth in South Africa in the short run but in the long run.
88
In our second model, we equally find that financial openness and economic growth in Nigeria have
a bidirectional causal relationship. This supports the finding of Hanh (2010). This evidence of
bidirectional causality between financial openness and economic growth suggests that financial
openness is necessary for enhancing economic growth. In turn, economic growth seems to be an
important condition for financial openness to take place and also thrive in Nigeria.
Finally, we find that past volatilities in financial openness and other macroeconomic variables in
our model did not lead to output volatility in Nigeria. This result supports the work of Kose et al
(2008).
5.2 Conclusion
This thesis focuses on the impact of financial openness on economic growth in Nigeria. The study
equally examines the direction of causality between financial openness and economic growth as
well as the impact of financial openness on output volatility in Nigeria. It uses two measures of
financial openness: de jure (Chin-Ito Index) based on Chinn and Ito (2012) and de facto capital
flows variables which are the sum of FDI, portfolio flows and other investments following
Aizenman (2004, 2008) and Aizenman and Noy (2009), for empirical analysis.
The thesis addresses three specified objectives namely; to examine the impact of financial
openness on economic growth in Nigeria; to empirically investigate the existence of a causal
relationship between financial openness and economic growth in Nigeria; and to determine the
impact of financial openness on output volatility in Nigeria. For the regression analysis, we use
bank and stock market data, international capital flow variables and institutional variables, among
others. These include Real GDP per capita, Credit to the private sector, Real Interest Rate, Human
Labour, Market Capitalization, Real Exchange Rate, and Institutional Quality Index. The study
applies the Autoregressive Distributed Lag Model based on unrestricted error correction model
(ARDL-UECM), Granger Causality Model and the Generalized Autoregressive Conditional
Heteroscedasticity (GARCH) Model to address the three objectives.
The results show positive impact of financial openness on economic growth in Nigeria both in the
short run and in the long run. Specifically we find that 1% increase in de facto financial openness
leads to an increase of 0.02% in economic growth in the short run and 0.26% in economic growth
89
in the long run. The results equally show that 1% increase in de jure financial openness leads to an
increase of 0.02% in economic growth in the short run and 0.26% in economic growth in the long
run. The results also reveal that credit to the private sector is negatively associated with growth,
indicating that there are problems with credit allocation and utilization in the country which could
be occasioned by weak regulation/supervision and non-adherence to prudential guidelines in the
financial system. We also find that real interest rate has a positive relationship with economic
growth. Although this is not very significant but the results support the McKinnon-Shaw
hypothesis, i.e. in the long run interest rate liberalisation will ultimately lead to rapid economic
growth. Human Labour (HML), is also statistically significant and positively related to Economic
growth in Nigeria. On the stock market side, the results show that market capitalization impacts
positively and significantly on economic growth. Our institutional quality variable which
represents governance and rule of law also shows some interesting results. In the short run it has a
negative relationship with economic growth but in the long run we see a positive relationship
between the two variables. This result attests to the fact that the present style of governance among
the leaders has serious negative impact on the growth of the Nigerian Economy. This is
exemplified by the fact that the principle of rule of law is not respected, corruption has been
enthroned in several leadership quarters, and there is no internal democracy even among the
political parties. When the quality of governance and institution is weak, it simply translates to
corruption, embezzlement of state funds meant for infrastructural development, and several other
anti -socio/economic outcomes.
The second regression result indicates the existence of bidirectional causality between financial
openness and economic growth in Nigeria. Even when we controlled for other macro variables,
the model results showed no significant difference.
The thesis further examines the impact of financial openness on output volatility in Nigeria. It
applies the Generalised Autoregressive Conditional Heteroskedasticity (GARCH) model. The
findings show that none of the two measures of financial openness contributed to output volatility
in Nigeria, within the period under review. Other control variables in the GARCH Model also
reveal a negative relationship with output volatility in Nigeria.
90
5.3 Policy implications of findings and Policy Recommendations
1. The first significant policy implication arising out of the empirical finding in the thesis is that
both the de facto and de jure measures of financial openness are equally robust and significantly
positively related to economic growth in Nigeria. Thus, policy makers or researchers who wish
to investigate long run or short run impact in the future could adopt either the de facto measure
of openness or the de jure measure of financial openness. This is because according to our model
results, the de facto and de jure financial openness measures showed similar impact in the long
run and short run respectively.
2. From this work, our knowledge of the various measurement issues associated with financial
openness has been enhanced and we have identified that both measures are potent and robust for
the Nigerian economy. Thus, we recommend that government should continue to remove barriers
to capital account transactions with every sense of objectivity, economic management dexterity
and in line with global best practices. Again, as earlier stated while justifying this study, an
important question requires policy-makers to decide, given that legal barriers have been
removed, how best to manage capital flows. Suffice it to note that countries such as China, India
and South Korea, are pointed out as poster child of success from openness. However, when
compared with some European countries, as well as many other developing countries like
Nigeria, these countries exhibit a much less “open” economy, measured either with trade and/or
financial openness indicators. Their high growth performance is more associated with a
“managed openness” policy, than a rush for “more openness” per se. In this light, having seen
the benefits of financial openness to the growth of the economy, the Nigerian Economic
managers should adopt the best economic management policies to guide international capital
flows and also ensure that the maximum benefits of such flows accrue to the country. In other
words, we recommend that the government and policy makers should adopt international best
practices and policies in guiding domestic financial system reforms and international capital
flows in order to ensure the maximum benefits of such policies to the economy.
3. Banks should be encouraged to extend more credit to the private sector. But there is a serious
need for discipline and discretion in credit allocation by the banks. Giving loans to friends and
91
cronies without serious certified profitable business ideas should be discouraged. Again the
government and financial sector players should educate the business community and other loan
seekers on the need to invest such credits in productive business ventures that will contribute to
rapid economic growth in the long run. To achieve this laudable goal, there is need to develop
and empower the relevant institutions. According to Prasad and Rajan (2008) “a successful
implementation of financial policy depends on the level of institutional and economic
development before the policy is implemented”. Ultimately, there is need to adopt value re-
orientation approach by the private sector towards banks’ borrowing and target investment in
productive activities of the economy in order to elicit economic growth (Orji, 2012).
4. As shown by our results, the liberalisation of interest rates is needed for generating higher
savings and investment in Nigeria. As it were, savings and investment can be facilitated by
maintaining higher real interest rates. Furthermore, monetary authorities and policy makers
should allow the market to determine the interest rates, but relevant policies must be put in place
to guard the market determined interest rates by setting objective margins for it. Again, sound
policies should be evolved to improve the efficiency of financial intermediaries while putting
inflationary pressures under control. This will ensure that lending and deposit rates put under
desirable levels. Depositors can be motivated to deposit more by increasing the deposit rates
while investors can be encouraged to use financial intermediaries by lowering the lending rates.
This is envisaged to improve the effect of financial sector development on economic growth
through real interest rate channel.
5. The government should enhance human capital development by developing the education sector.
Having seen from our empirical analysis that the quality of Human labour and capital contributes
to growth, policy makers should evolve sound education policies that will help in enhancing the
capacity of our teaming youths to contribute positively to the growth process of the nation.
Quality education with sufficient funding should be emphasized at all levels of government. This
is vital because management of financial openness and all international capital flows that will
contribute positively to the growth of the economy can only be accomplished by educated sound
92
minds. In this regard, we strongly advocate a sustainable progressive increase in budgetary
allocation to the education sector to 26% in Nigeria by the year 2020 and beyond.
6. Also, policy makers and monetary authorities should ensure that capital markets in Nigeria are
strategically developed and repositioned such that they are incorporated and integrated into the
financial system and the economy as a whole. The results indicate that the level of market
capitalization in Nigeria is positively related to growth. Thus, there is need to continue the drive
towards maximising the economic growth potentials of the Nigerian Stock Markets by
adequately ensuring that they keep providing funds to investors for long term investment,
business and development projects. As noted by Adjasi and Biekpe (2006), the efficiency and
productivity effects of stock market on economic growth are strong and positive when stock
markets are liquid and active. The recent political impasse between the Securities and Exchange
Commission (SEC) and the National Assembly in Nigeria whereby the SEC is being starved of
budgetary funds for their statutory operations should be highly discouraged.
7. The country’s institutional quality should be comprehensively reviewed and upgraded. Strong
emphasis should be made on deepening the country’s democracy, reforming the governance and
electoral systems and reorganizing the socio/political structures in the country. Respect for the
rule of law should be given priority by the leaders and the led. All avenues through which
corruption is encouraged in the system should be discontinued and if anybody is found guilty of
any corrupt practice, the person should be made to face the full wrath of the law. This is because
according to our finding, poor governance, which is exemplified by corruption and lack of
respect for the rule of law are detrimental to growth. However, if these anomalies are corrected,
improved institutional quality will impact positively on growth in the long run. To ensure
compliance and achieve maximum result, the judiciary and various anti-corruption agencies
should be properly funded and given full independence to function properly. This will enable
them to deal with cases of corruption and other governance issues decisively no matter whose
ox is gored.
93
8. This thesis supports a-bidirectional causality between financial openness and Economic growth
in Nigeria. We therefore recommend that, on one hand, the government and policy makers
should make suitable policies for attracting foreign capital flows along with deepening the
domestic financial system in order to enhance growth. On the other hand, the country’s growth
should be adequately managed in order to generate more gains not only in terms of domestic
financial development but also in terms of financial openness by attracting productive
international capital inflows. The recent rebasing of Nigeria’s GDP should not be an end in itself
but efforts should be geared towards formulating useful economic policies that will lead to
sustained economic growth in the Nigeria.
9. Emphasis should be placed on more robust domestic economic structural reforms such as
promoting a competitive and viable domestic banking system, with adequate regulatory and
supervisory framework. This should be complemented by other macroeconomic stability
policies. That is, fiscal deficits, rapidly depreciating exchange rate and high inflation should be
put in check. This is one of the ways to ensure that financial openness continues to contribute to
growth while lowering output volatility.
5.4 Research Recommendations for Further Studies
This study has considered the impact of financial openness on economic growth and output
volatility using the de jure and de facto measures. However, the following areas are
recommended for further research.
(i) Studying the impact of financial openness on other key macroeconomic variables like
Investment, Government Expenditure, Taxation, Imports and Exports using the Chinn-Ito
index.
(ii) Analyzing the impact of financial openness on regional economic blocks like ECOWAS,
WAEMU, WAMZ, among others, using the de jure and de facto measures of financial
openness.
(iii) Investigating the effect of financial openness policies on the economy through various
other sectors of the economy using the Chinn-Ito Index or the de facto measure.
94
REFERENCES
Agenor, P. (2003), “Benefits and Costs of International Financial Integration: Theory and Facts” The World Economy (26) 8: 1089-1118
Agenor, P. and Aizenman, J. (1999), “Volatility and the Welfare Costs of Financial Market
Integration”, in the Asian Financial Crisis; Causes, Contagion and Consequences, ed. By Pierre-Richard Agenor, Marcus Miller, David Vines and Axel Weber, Cambridge: Cambridge University Press.
Agu, C.C, Orji, A. And Eigbiremolen G. (2014) “Financial Liberalization, Interest Rate Structure
and Savings Mobilization: The Nigerian Experience” International Journal of Current Research 6 (1), January, 2014
Agu, C.C. (1988), “Interest Rate Policy in Nigeria: An Attendant Distortion” Savings and
Investment Journal, 12 (1). Aguiar, M. and Gopinath, G. (2007), “Emerging Market Business Cycles: The Cycle is the Trend”
NBER Working Paper 10734 Ahmed, M. and Suardi, S (2009), “Do Financial and Trade Liberalization Cause Macroeconomic
Volatility in Sub-Saharan Africa?” Centre for Strategic Economic Studies, Working Paper No. 44, Victoria University, Melbourne
Aizenman, J. and Noy, I. (2009), “Endogenous Financial and Trade Openness”. Review of
Development Economics, 13: 175-189 Aizenman, J. and Noy, I. (2004), “Endogenous Financial Openness: Efficiency and Political
Economy Considerations”, NBER Working Paper No. W10144. Aizenman, J., (2008). "On the hidden links between financial and trade opening," Journal of
International Money and Finance, Elsevier, 27(3): 372-386 Akpan, H. E (2013) “Growth without Development in Nigeria: Confronting the Paradox”, Paper
presented at the 43rd Annual ICAN Conference held at the International Conference Centre, Abuja, 8 – 10, October, 2013
Alfaro, L., Chanda, A., Kalemli-Ozcan, A., and Sayek, S. (2004), “FDI and Economic Growth:
the Role of Local Financial Markets,” Journal of International Economics 64: 89-112 Aremu, J.A (2005), “Attracting and Negotiating Foreign Direct Investment with Transnational
Corporations in Nigeria.” CBN Papers
95
Babajide, F. (2008), “Financial Liberalization Policies and Economic Growth: Panel Data Evidence from Sub-Saharan Africa” African Development Review (20) 3: 549-574 Backus, D.K., Kehoe, P., Kydland, F., (1992) “International Real Business Cycles”, Journal of
Political Economy 100 (4): 745-775. Bahmani-Oskooee, M and and Brooks, T.J (1999), “Bilatera J-curve between U.S and Her Trading
Partners”, Weltwirtschaftliches Archiv, (135): 156-165 Banerjee A. (1993), “Co-integration, Error Correction, and the Econometric Analysis of Non-
Stationary Data”. Oxford, Oxford University Press. Bardsen, G. (1989). Estimation of Long-Run Coefficients in Error Correction Models, Oxford
Bulletin of Economics and Statistics, 51: 345-350. Barro, R.J. and Xavie S. (1995) “Economic Growth”, New York: McGraw Hill. Batini N. (2004) “Achieving and Maintaining Price Stability in Nigeria” IMF Working Paper
WP/04/97 Baxter, M and Jermann, U.J (1997), “The International Diversification Puzzle is worse than you
think.” The American Economic Review, 87(1): 170-180.
Bekaert G., Harvey, C.R. and Lundblad, C. (2006). “Growth Volatility and financial liberalization,” Journal of International Money and Finance, 25: 370-403.
Bekaert G., Harvey, C.R. and Lundblad, C. (2005). “Does Financial Liberalization spur Growth”, Journal of Financial Economics 77(1):3-55.
Bhagwati, J. (1998), “The Capital Myth: The difference between Trade in Widgets and Trade in
Dollars” Foreign Affairs, (77): 7-12. Bollerslev, T. (1986), “Generalized Autoregressive Conditional Heteroskedasticity”, Journal of
Econometrics, 31, 307-327. Buch C.M., Jorg, D and Christian P. (2005). “Financial Openness and Business Cycle Volatility”:
Kiel Institute for World Economics. Working paper, Germany. Buch C.M., and Yener, S. (2005). “Consumption volatility and financial openness” Kiel working
Paper 1260 Braun, M. and Raddatz, C. (2007), “Trade Liberalization, Capital Account Liberalization and the
Real Effects of Financial Development,” Journal of International Money and Finance, 26(5): 730-761
96
Calderon, C. and Yeyati, E.L (2007), “Financial and the Poor: What matters and What Doesn’t?’ The World Bank, manuscript
Campos, N.F and Coricelli, F. (2012), “Financial Liberalization and Reversals; Political and
Economic Determinants” Economic Policy fifty-fourth Panel Meeting Hosted by the National Bank of Portland Warsaw, 27-28 October 2011
Caprio, G. and Klingebiel, F (1995), “Bank Insolvencies: Cross Country Experience”, World Bank
Working Papers No. 1620 Carmignani, F and Chowdhury, A (2005) “The Impact of Financial Openness on Economic
Integration: Evidence from the Europe and the Cis” Department Of Economics, University Of Milan – Bicocca Working Paper Series
Central Bank of Nigeria (2012) Statistical Bulletin Central Bank of Nigeria (2010) Statistical Bulletin Central Bank of Nigeria (2005) Statistical Bulletin Central Bank of Nigeria (2004) Statistical Bulletin Chanda, A. (2005), “The Influence of Capital Controls on Long Run Growth: Where and How
Much?” Journal of Development, 77 (2): 441-466 Chanda, A. (2001). “The Influence of Capital Controls on Long Run Growth: Where and How
Much?” Brown University, Mimeo Chinn, M.D. and Ito, H. (2012), “Notes on the Chinn-Ito Financial Openness Index 2010 Update”
NBER Working Papers Chinn, M.D and Ito, H. (2006), “What Matters for Financial Development? Capital Controls,
Institutions, and Interactions,” Journal of Development Economics 81 (1); 163-192. Chinn, M.D and H. (2002) capital account liberalization, institutions and financial development:
Cross country evidence.” NBER Working Paper No. 8967 Chowdhury, A., and Mavrotas, G. (2006), FDI and Growth: What Causes What? World economy,
29(1), 9-19. Coricelli, F., Jazbec, B., Masten, I., (2008), “Sources and Obstacles for Growth in Transition
Economies; the Role of Credit”, festschrift in Honour of G.Calvo, MIT Press Dermirguc-Kunt, A. and Detragiache, E. (1998), “Financial Liberalization and Financial Fragility”
IMF Working Paper WP/98/83 Diercks, A. (2012), “Capital Account Liberalization: A New look” NBER Working Papers
97
Easterly, W., Islam, R., and Stiglitz, J., (2001). Volatility and Macroeconomic Paradigms for Rich and Poor Countries. In: Dreze, J. (Ed.), Advances in Macroeconomic Theory. Palgrave, New York.
Edison, H.J; Levine, R. l and Slok, T. (2002), “International Financial Integration and Economic
Growth,” Journal of International Money and Finance, 21 (60):749-776 Edwards, S. (2001), “Capital Mobility and Economic Performance: Are Emerging Economies
Different?” NBER Working Paper No.8076 Eichengreen, B. and Leblang, D. (2003), “Capital Account Liberalization and Growth: Was Mr.
Mahathir right”? International Journal of Finance and Economics (8):20 -224 Engle, R. F. (1982), “Autoregressive Conditional Heteroscedasticity with Estimates of the
Variance of United Kingdom Inflation”, Econometrica, 50, 987-1007 Fratzscher, M and Bussiere, M (2004), “Financial Openness and Growth: Short- Run Gain, Long-
Run Pain? European Central Bank Working Paper Series No. 348 French, K.R and Poterba J.M (1991), “Investor Diversification and International Equity Markets”
NBER Working Paper no 3609 Fry, M.J. (1997), “In Favour of Financial Liberalization”, Economic Journal, 107: 442-754 Gallego, F and Loayza (2002) “The Golden Period for Growth in Chile- Explainations and
Forecast.” Journal of Econonomics Chilena, 5:37-67 Ghatak, S. (1997). “Financial Liberalisation: The Case of Sri Lanka”, Empirical Economics,
22:117-129. Giovanni, J.D and Levchenko, A.A (2006), “Trade Openness and Volatility” IMF Development
Studies Working Papers No.219 Greenaway, D, Morgan, W and Wright, P. (2002), “Trade Liberalization and Growth in
Developing Countries” Journal of Development Economics, 67; 229-244 Grilli, V. and Millessi-Ferretti, G.M (1995), “Economic Effects and Structural Determinants of
Capital Controls”, IMF Staff Papers, (42) 3: 517-550 Gupta, R. and Komen, K. (2008) “Time Aggregation and the Contradictions with Causal
Relationships: Can Economic Theory Come to the Rescue?” University of Pretoria Department of Economics Working Paper Series 2008-02 February 2008
Hagen, J.V and Zhang, H (2006), “Financial Liberalization in a Small Open Economy” CESifo
Working Paper Series 1771, CESifo Group Munich
98
Henry, P.B and Lorentzen, P.L (2003), “Domestic Capital Market Reform and Access to Global Finance: Making Markets Work,” Research Papers 1820, Standford University, Graduate, Graduate School of Business.
Henry, P.B (2006), “Capital Account Liberalization: Theory, Evidence and Speculation” NBER
Working Paper 12698 Hermes, N. and Lensik, R. (2005), “Does financial liberalization influence saving, investment and
economic growth? Evidence from 25 emerging market economies; 1973-1996” UNU-WIDER Research Paper No.2005/69, Finland: United Nations University.
Honohan, P. (2000), “How Interest Rates Changed Under Financial Liberalization- a Cross
Country Review,” Policy Research Working Series 2313 Johansen, S. (1995), “Likelihood-Based Inference in Co-integrated Vector Autoregressive
Models”. Oxford, Oxford University Press. Johansen, S. (1991), “Estimation and Hypothesis Testing of Co-integration Vectors in Gaussian
Vector Autoregressive Models,” Econometrica 59:1551-1580 Johansen, S.and Juselius, K. (1990), “Maximum Likelihood Estimation and Inference on Co-
integration- with Applications to the Demand for money”. Oxford Bulletin of Economics and Statistics 52: 169-210
Johansen, S. (1988) “Statistical Analysis of Co-integration Vectors”, Journal of economic
Dynamics and Control, 12: 231-254 Kama, U (2006), “Recent Reforms in the Nigerian Banking Industry: Issues and Challenges” CBN
Bullion, 30:3 Kaminsky, G.L and Schmukler, S.L (2003), “Short-Run Pain, Long-Run Gain: The Effects of
Financial Liberalization” IMF Working Paper WP/03/34 Kaminsky, G.L and Reinhart, C.M (1999), “The Twin Crises: The Causes of Banking and Balance
of Payments Problems,” American Economic Review 89 (3): 473-500 Klien, M.W and Olivei, G.P (2008), “Capital Account Liberalization, Financial Depth and
Economic Growth” Journal of International Money and Finance, 27(6): 861-875. Klien, M.W (2003), “Capital Account Openness and the Variety of Growth Experience” NBER
Working Papers 9500 Klien, M.W and Olivei, G (2001), “Capital Account Liberalization, Financial Depth and Economic
Growth” Mimeo, Tufts University
99
Klien, M.W and Olivei, G (1999), “Capital Account Openness and the Variety of Growth Experience” NBER Working Paper 7384
Kose, A., Ayhan, M., Prasad, E, Rogoff, K. Shang-Jin, W., (2008), “Financial Globalization: A
Reappraisal”, IMF Staff Papers Kose, A. Prasad, E., Terrones, M.E (2003), “Financial Integration and Macroeconomic Volatility”
IMF Staff Papers 50: 119-142 Kose, A (2002), “Explaining Business Cycles in Small Open Economies: How much do World
Prices Matter”? Journal of International Economies 56:299-327 Kraay, A (1998), “In Search of the Macroeconomic Effects of Capital Account Liberalization”
Manuscript, the World Bank Lal, D (2012) “Is the Washington Consensus Dead?” Cato Journal, 32 (3): 493-512 Lane, P.R., and Milesi- Ferretti, G.M (2007), “The External Wealth of Nations Mark II: Revised
and Extended Estimates of Foreign Assets and Liabilities, 1970-2004” Journal of International Economics 73 (2): 223-250
Lane, P. R and Milesi-Ferretti, G.M (2005), “A Global Perspective on External Positions,” NBER
Working 11589 Law (2006), “The Determinants of Financial Development from Openness and Institutional
Quality Perspectives” Working Papers, Department of Economics, Universiti Putra Malaysia
Levine, R. (2001), “International Financial Liberalization and Economic Growth”. Review of
International Economics 9(4): 688-702 Lewis, K.K (1999), “Trying to Explain Home Bias in Equities and Consumption” Journal of
Economic Literature, 3792): 571-608 Loayza, N.V and Raddatz C. (2007), “The Structural Determinants of External Vulnerability” The
World Bank Economic Review, 21 (3): 359-387 Loayza, N.V and Ranciere, R. (2006), “Financial Development, Financial Fragility and Growth”
Journal of Money, Credit and Banking, 38(4): 1051-1076 Luintel, K.B (2008), “Financial Structure and Economic Growth”, Journal of Development
Economics, 86 (1), 181-200 Mailafia, O. (2006), “Capital Account Liberalization: Reflections on Theory and Policy” CBN
Economic and Financial Review 44 (4)
100
Maniragaba,V.N (2011), "Effects of Financial Liberalisation on Money Demand Economic Growth: Uganda’s Experience (1978-2008)". Makerere University, Uganda.
Masron, T.A (2009), “Macroeconomic Synchronization and Foreign Direct Investment Inflow into ASEAN: Evidence from Malaysia’s Experiences” International Business Research, 20 (3): 11-46
Mckinnon, R.I (1973), “Money and Capital in Economic Development”. Washington D.C, The
Brookings Institution Mishkin, F.S (2006), “The Next Great Globalization: How Disadvantaged Nations can Harness
their Financial System to Get Rich” Princeton, NJ: Princeton University Press Montiel, P.J (1995), “Financial Policies and Economic Growth: Theory, Evidence and Country-
Experience from Sub-Saharan Africa”, AERC Special Paper no. 18, Nairobi. Narayan, P.K and Narayan, S. (2005), "Estimating Income and Price Elasticities of Imports for
Fiji in a co-integration framework” Economic Modelling 22: 423-438 Neumann, R. and Ron P. (2008) "Volatile capital flows: Interactions between de jure and de facto
financial liberalization." Economics Bulletin, 6 (2): 1-10 Nigeria Bureau of Statistics Report 2012 Obiechina, M.E (2010) “Capital Flows and Financial Crises: Policy Issues and. Challenges for
Nigeria” Economic and Financial Review 48 (1): 193-112. Odhiambo, N. M (2009), "Interest Rate Liberalisation, Financial Deepening and Economic
Growth in South Africa": Ninth Annual IBER & TLC Conference Proceedings 2009 Odhiambo, N.M. (2009b). "Interest Rate Reforms, Financial Deepening And Economic
Development In Kenya: An Empirical Investigation", The Journal of Developing Areas, 43 (1): 295-313
Odhiambo, NM. (2009c)." Interest rate liberalization and Economic Growth in Zambia: A
Dynamic linkage", African Development Review, 21(3): 541 - 557. O’Donnell, B. (2001), “Financial Openness and Economic Performance”, (unpublished cross-
country research work, Dublin, Ireland: Trinity College). Ogbonna, I.C, Uwajumogu, N.R, Nwokoye E, Nzeribe G (2012) “Economic Growth and Foreign
Direct Investment in Nigeria: An Empirical Investigation Journal of Economics and Sustainable Development 3(13): 34-40
Ohiorhenuam, J.F.E. (2003) “Development Thought, Policy Advice and Economic Development
in Africa in the 20th Century”: Evolution of Development Thought: 1-38
101
Ojo, M. O (2006), “Capital Account Liberalization: The ECOWAS Experience”, CBN Economic and Financial Review 44:4
Okpara, G.C (2010), “The Effect of Financial Liberalization on Selected Macroeconomic
variables: Lessons from Nigeria”. The International Journal of Applied Economics and Finance 4 (2): 53-61
Omoke (2010) “The Causal Relationship among Financial Development, Trade Openness and
Economic Growth in Nigeria”. International Journal of Economics and Finance, 2(2):1-15
Onosode, G.O. (1997) Foreign Investment Not a Panacea for Resource Mobilization: Daily
Times: March 5. Orji, A, Eigbiremolen, G and Ogbuabor, J.E. (2014) “Financial Liberalization and Private
Investment: Further Evidence from Nigerian Data”, Review of Economics & Finance 4 (2): 77-86
Orji, A., Onyia A.M and Ani, E.G (2013) “Calming the Tide: Private Investment and
Macroeconomic Uncertainty in Nigeria” Asian Journal of Empirical Research, 2013, 3(5):633-653
Orji, A. (2012) “Bank Savings and Bank Credits in Nigeria: Determinants and Impact on
Economic Growth”. International Journal of Economics and Financial Issues, 2 (3): 357-372
Osinubi,T.S and Amaghionyeodiwe L.A (2010) “Foreign Private Investment And Economic
Growth In Nigeria” Review Of Economic And Business Studies 3(1):105-127 Owusu, E.l (2012), “Financial Liberalization and Economic Growth in ECOWAS Countries”: An
Unpublished PHD Research Work. Oyatoye, E. O., Arogundade, K. K, Adebisi, S. O and Oluwakayode, E. F. (2011) “Foreign Direct
Investment, Export and Economic Growth in Nigeria” European Journal of Humanities and Social Sciences 2 (1):67-86
Özdemir, D and Erbil, C. (2008), “Does Financial Liberalization Trigger long-run Economic
Growth? Evidence from Turkey and Other Recent EU Members” EcoMod International Conference on Policy Modeling, Berlin, Germany, July 2-4, 2008
Pagano. M. (1993), “Financial Market and Growth: An Overview’, European Economic Review,
37 (1993): 613 – 622. Pesaran, M.H, Shin, Y and Smith, R.J (2001), “Bounds Testing Approaches to the Analysis of
Level Relationships”, Journal of Applied Econometrics 16 (3): 289-326
102
Pesaran, M.H and Pesaran, B. (1997), Working with Microfit 4.0: Interactive Econometric Analysis. Oxford, Oxford University Press.
Pesaran, M.H, Shin, Y and Smith, R.J (1996), “Testing the Existence of Long Run Relationship”
DAE Working Paper Series No.9622, Department of Economics, University of Cambridge.
Pesaran, M.H and Shin, Y (1995), “Autoregressive Distributed Lag Modelling Approach to Co-
integration Analysis”. DAE Working Paper Series No.9514, Department of Economics, University of Cambridge
Poon, W.C (2010), “A Monetary Policy Rule: The Augmented Monetary Conditions Index for
Philipines Using UECM and Bounds Tests”, Discussion Paper 04/10, Monash University, Department of Economics.
Portes, R. and Rey, H (2003), “The Determinants of Cross Border Equity Flows” Manuscript
Princeton University. Prasad, E., Rajan, R. and Subramanian, A. (2007), “Foreign Capital and Economic Growth”
Brooking Papers on Economic Activity, 2007 (1): 153-209 Prasad, E., Rogoff, K., Wei, S.J and Kose, M.A. (2003), “The Effects of Financial Globalization
on Developing Countries: Some Empirical Evidence. IMF Occasional Paper No.220. Washington, D.C
Prati, A and Tressel, T (2006), “Aid Volatility and Dutch Disease: Is there a Role for
Macroeconomic Policies”? IMF Working Paper WP/06/145 Quinn, D.P and Toyoda, A.M (2007), “Ideology and Voter Sentiment as Determinants of Financial
Globalization.” American Journal of Political Science 51: 344-363 Quinn, D. (1997), “The Correlates of Change in International Financial Regulation”, American
Political Science Review 91(3):531-551. Quinn, D. and Inclan, C. (1997), “The Origins of Financial Openness: A Study of Current and
Capital Account Liberalization”, American Journal of Political Science 41(3): 771-813. Ramey, G and Ramey, V.A (1994), “Cross-Country Evidence on the Link between Volatility and
Growth,” NBER Working Papers 4959 Raddatz, C., (2007). “Liquidity needs and vulnerability to Financial Underdevelopment”, Journal
of Financial Economics 80 (3): 677-722 Razin, A. and Rose, A. (1994), “Business Cycle Volatility and Openness: An Exploratory Cross-
Section Analysis.” NBER Working Papers 4208
103
Rodrick D. (1998), “Who Needs Capital Account Convertibility?” in Stanley Fisher, et al., Should the IMF Pursue Capital Account Convertibility? Essays in International Finance, No 207, International Finance Section, Department of Economics, Princeton University, Princeton N.J.
Serven, L. and Schmidt-Hebbel, K. (2002), “Financial Liberalization, Saving and Growth” Paper
Presented at the Banco de Mexico Conference on Macroeconomic Stability, Financial and Economic Development”, Mexico, DF, Nov.12-13
Shaw, E.S, (1973), “Financial Deepening in Economic Growth” New York: Oxford University
Press Shrestha, M.B and Chowdury, K (2007), “Testing Financial Liberalization Hypothesis with ARDL
modelling approach”, Applied Financial Economics 17(18): 1529-1540 Solow, R. M (1956), “A Contribution to the Theory of Economic Growth” Quarterly Journal of
Economics 70 (1): 65-94 Stiglitz, J.E (2002). Globalization’s Discontents. New York: W.W Norton and Company Stiglitz, J.E (1998), “Towards a New International Economic Architecture: Principles and
Policies.” Presented at the European Investment Bank, Fiesole Italy, October, 15. Tang, T. C. (2003). “Japanese aggregate import demand function: Reassessment from the ‘bound’ testing approach. Japan and the World Economy. 15 (4): 419-436. Tesar, L.L and Werner, I. M (1995), “Home Bias and High Turnover” Journal of International
Money and Finance. 14 (4): 467-492 Toda, H. Y. and Yamamoto, T. (1995), “Statistical Inference in Vector Autoregressions with
Possibly Integrated Processes” Journal of Econometrics, 66, 225-250. Uchendu, O.A. (1993), “Interest Rate Policy in Savings and Investment in Nigeria” CBN Economic
and Financial Review 31:13-25 Udah, E.B (2010), “Macroeconomic Reforms, Government Size and Investment Behaviour in
Nigeria: An Empirical Investigation” Journal of Sustainable Development in Africa. 12 (1):258-280
Williamson, J. (1989), “What Washington Means by Policy Reform” In J.Williamson (ed.) Latin
American Re-adjustment: How Much Has Happened? Washington: Institute for International Economics.
World Bank Development Indicators 2012 Zhang, K.H.L., (2001). “Does Foreign Direct Investment Promote Economic Growth? Evidence
from East Asia and Latin America”. Contemporary Economic Policy 19: 175-185.
104
APPENDIX
A. Selected Macroeconomic Variables
-8.754-10.752
7.5436.467
12.766
-0.6180.4342.090.91-0.307
4.9942.8022.716
0.474
5.3188.164
21.177
10.33510.585
5.3936.2116.9725.9846.968.724
7.4 7.1
-15
-10
-5
0
5
10
15
20
25
Source: World Bank and CBN (2012)
Fig. 1 NIGERIA GDP GROWTH RATE
105
11.63228745
-24.06710072
-3.92085612
-16.57844108
16.92732415
-0.110733084
-32.05731051
-13.74923058
-5.702380374
-22.91094097
-12.46367043
16.2132807
25.13000989
7.127715426
-12.22751594
11.46917384
-5.098434484
8.560264468
-1.281265538-1.513291538-2.223439769
11.57313977
4.052791229
23.82479618
-7.252417692
13.36478135
-40
-30
-20
-10
0
10
20
3019
86
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Source : World Bank and CBN (2012)
Fig.2 Nigeria:Real Interest Rate
6.2511.765
34.211
49.02
7.89512.195
44.565
57.14357.416
72.729
29.292
10.6737.8626.6186.938
18.869
12.88314.03315.00117.856
8.2185.413
11.58112.54313.7210.812.1
0
10
20
30
40
50
60
70
80
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Source : World Bank and CBN (2012)
Fig. 3 Nigeria: INFLATION RATE
106
B: The Normality Test
0
5
10
15
20
25
-1 0 1 2
S eries: S tandardized ResidualsS ample 1986:2 2011:4Observations 103
Mean 0.061664Median -0.065751Maximum 2.560122Minimum -1.642017S td. Dev. 0.825837S kewness 0.469314K urtosis 3.531363
Jarque-B era 4.992796P robability 0.082381