# Macroeconomic Financial Leverage And Financial Crisis In Nigeria Finance Essay

This study consistutes a first empirical analysis on the relationship between Financial Leverage in the Macro economy, Global Financial Crisis and Oil Price Volatility for the Nigerian Economy employing the Financial Soundness Indicators as developed in 1998 by the International Monetary Fund, World Bank and European Central Bank for Country-Specific and External Factors for the period 1970 to 2010. By employing current data, this study carries out, Cointegration Technique, The Vector Error Correction Mechanism (VECM), Granger Causality, and Exponential Generalized Autoregressive Conditional Heteroscedascity (EGARCH) methodology. The results provide evidence of an equilibrium relationship between Macroeconomic Financial Leverage and the Financial Soundness Indicators with the discrepancy between the short run and equilibrium value being corrected quarterly. This study also provides evidence of significant impact of Financial Soundness Indicators both country specific and external factors on Macroeconomic Financial Leverage. A main innovation is that various type of tests are carried out, sub samples are also examined to judge the robustness of results. Across, sub samples and full sample period this study provides evidence of asymmetry and persistence of shocks on volatility of oil price. This study contributes to literature in the following ways; by focusing on macoreconomic financial leverage in an emerging market economy, and examining the long run, short run dynamics and direction of influence of country specific and external factors , and examining the impact of the Global financial crisis on macroeconomic financial leverage, also modelling oil price volatility employing current daily data in the Generalized Autoregressive Conditional Heteroscedasticity (GARCH) framework and examining the impact of the Global Financial crisis on the asymmetrics and persistency of oil price shocks to volatility, furthermore. This study is important to policy makers in financial planning and control, investors, in making optimal portfolio decisions and international institutions like International Monetary Fund in analysing the Nigerian economy. This study also creates a background for studying similar issues in countries such as Algeria, Venezuela, and Egypt. This study can also be used for comparative analysis for other Organization of Petroleum Exporting Countries (OPEC) member countries. This study recommends the creation of new financial architecture, greater diversification of the economy and the export market.

## CHAPTER ONE

## 1.1 INTRODUCTION: BACKGROUND OF STUDY

Nigeria as a country has certain features that make it important and interesting to be studied. It is an Emerging Market Economy, according to the Emerging markets Database of the International Finance Corporation of the World Bank Group and one of their fastest-growing client countries; it is already the ninth-largest in their portfolio and has the largest exposure in Africa. Heavily dependent on Oil Revenue, it is Africa’s leading producer and exporter of oil, with the largest natural gas reserves. With the largest population in Africa, of over 148million people [1] , it is also ranked as the 7th Largest Exporter of Oil in the world, and a prominent member of the Organization of Petroleum Exporting Countries (OPEC). It is argued that in view of the fact that United States is the major importer of Nigerian Oil, any problem that emanates from the US economy, affects the Nigerian economy. Nigeria is also popularly known to be in debt both domestic and external, major external debt are funded by the World Bank and African Development Bank.

This study contributes to existing knowledge and literature by providing evidence of an existing empirical relationship among these unique features of the Nigerian economy, by examining Macroeconomic financial leverage in the form of treasury bills issued by the Government, Domestic and External Debt. The Global financial crisis, its impact on the Nigerian Economy and Oil price Volatility is also examined.

Elder and Serletis (2009), found empirical evidence that uncertainty about oil prices tend to depress investment in the United States. Rafiq, Salim and Bloch (2008), argued that studies analysing the impact of oil price volatility are needed for emerging market economies. Treasury Bills [2] were the first money market instruments issued by the Central Bank of Nigeria (CBN), in April 1960, other forms of financing are domestic and external debt. Financial Leverage is important in determining financial risk. In the case of macroeconomic financial risk, macroeconomic financial leverage [3] is measured by interest expense, which causes variability in the Gross Domestic and National Product.

In 1970, the end of the Biafran [4] War coincided with a rise in the world oil price, and yielded instant wealth for the Nigerian economy. It is argued that the current global crisis is the worse and deepest recession in the history of the world, because oil is the lifeblood of modern economics [5] , the recession is accompanied by a significant volatility in the price of crude oil. A good example of the relevance of oil in world economies is the recent Gulf of Mexico Oil Spill which is estimated to cost $32 billion of British Petroleum’s assets, and has lead to a loss of $18billion. The Oil Spill also led to the resignation of the British Petroleum boss Tony Hayward. It is widely known that oil as a resource is a main source of revenue in Nigeria therefore any study and analysis on the Nigerian Finance and Investment sectors cannot ignore the Oil Dependency factor.

A phenomenal crash of the Wall Street occurred in 1929 [6] , in which share prices on the New York Stock Exchange collapsed on the 29th of October, 1929 popularly known as the Black Tuesday [7] . This lead to an economic depression that affected Western industrialized countries, aside from a vicious cycle of bank and business failures, increase in unemployment, oil price shocks also increased as a result of the crash. Therefore, the relationship between oil prices, financial markets, and the economy has been and continues to be an enthusiastic subject to researchers such as Hamilton (1983, 1988), Chen, Roll, Ross, and Nai-Fu (1986), Ferderer (1996), Uri (1996), Rotemberg and Woodford (1996), Cunado and Gracia (2003), Sardosky and Basher (2006), Cologni and Manera (2007), Rafiq, Salim and Bloch (2008), Faroq & Bradley (2009), Rano Aliyu (2009) and (Oskooe (2010).

Oil plays a significant role in economic activity, as such there is a growing body of literature on the relationship between financial markets, oil price volatility and economic activities. It is argued that the oil industry is very important, and a lot depends on the price of oil, researchers have examined how oil prices and its volatility affect major financial markets in the world, such as emerging markets (Sadorsky & Basher (2006). The Figure 1.1 [8] shows the dates of substantial change in the price of oil in Nigeria. A common situation with the Nigerian Oil industry is pipeline vandalism, kidnappings, militant takeover and worker’s strike. Oil was discovered in 1956, with major exports beginning in 1958, and oil revenue became a dominant factor in the Nigerian economy as reported by the World Bank [9] .

Economic growth and development in Nigeria, continues to be closely linked to oil. The Figure 1.2 [10] demonstrates the dependence of the economy on oil. Therefore, any relevant study on the finance and investment sectors of the Nigerian economy, examines the impact of oil price volatility. It can be argued that economic performance and growth of the finance and investment sectors in Nigeria follows the direction of oil production, exploration, and exportation.

## 1.2 STATEMENT OF RESEARCH PROBLEM

The Global Financial Crisis which is argued to be as a result of the credit crunch in the United States mortgage market continues to extend to the economies of the world majorly through international trade. Its impact on the Nigerian economy is more evidenced with the major swings in the price of Oil, Government revenue, and heavy debt by the Government. It is generally accepted that fluctuations in the price of oil will certainly affect oil exporting and importing countries, with Nigeria as an emerging market economy dominated by oil revenue inclusive. Therefore volatility in the price of oil, its causes and effects is important in the financial, investment and economic sectors of the Nigerian economy, by increasing uncertainty whether to invest or not and the future return on investment.

Therefore, in view of the above problems, the following questions are raised:

To what extent does the global financial crisis affect the investment and finance sectors of the Nigerian Economy, particularly macroeconomic financial leverage?.

To what extent does economic and financial indicators affect the finance and investment sector of the Nigerian Economy, particularly macroeconomic financial leverage?.

To what extent does the global financial crisis affect oil price volatility in Nigeria?.

To what extent do shocks to oil price volatility have asymmetric and persistent effects?.

## 1.3 AIMS AND OBJECTIVES

This study has three main objectives

Analysing the volatile nature of the Nigerian High Grade Bonny Light oil price to examine the asymmetric and persistency of shocks to volatility and examining

Its relationship on Macroeconomic Financial leverage during the period under study, by examining the impact of Government oil revenue.

Analysing the impact of the financial crisis on macroeconomic financial leverage by employing the Financial Soundness Indicators, commonly known as FSIs developed by the International Monetary Fund, World Bank and European Central Bank (country specific and external factor variables).

## 1.4 SIGNIFICANCE OF THE STUDY

Elder and Serletis (2009), found empirical evidence that uncertainty about oil prices tend to depress investment in the United States. Rafiq, Salim and Bloch (2008), argued that studies analysing the impact of oil price volatility are needed for emerging market economies. Therefore, this study contributes to existing literature on emerging market economies. The debate on whether oil as a resource is a curse rather than a blessing to the Nigerian economy is a standard argument, because of the overdependence of the economy on oil revenue hence economic activities moves in the direction of oil price changes. Therefore, oil price shocks affect the various sectors of the economy more significantly the investment and finance sectors.

This study consistutes a first empirical analysis on examining the relationship between Macroeconomic Financial Leverage, Global Financial Crisis and Oil Price Volatility for the Nigerian Economy employing the Financial Soundness Indicators as developed in 1998 by the International Monetary Fund, World Bank and European Central Bank. The global financial crisis has been argued to affect the Investment and Finance sector through the oil industry, as a result of it’s over dependence on oil revenue for stability. This research aims at investigating theoretically and empirically the relationship between macroeconomic financial leverage, oil price volatility and global financial crisis by employing secondary data from the period 1970-2010.

In order to analyse the long and short run equilibrium relationships and direction of influence of macroeconomic financial leverage variables, and financial soundness indicators, the Cointegration Technique, Vector Error Correction model (VECM) and Granger Causality Tests are employed. The variables used in this study, are based on the Financial Soundness Indicators developed by the International Monetary Fund with the support of other international organizations such as the World Bank, and European Central Bank in 1998.

Following the approach of Narayan and Narayan (2007) of modelling oil price volatility, the Exponential Generalized Autoregressive Conditional Heteroscedasticity Model (EGARCH) is employed to model the volatility of BonnyLight High Grade Nigerian Oil, for the purpose of investigating asymmetrics and persistency of oil shocks. Similarly, this research employs current daily data of oil prices as opposed to quarterly data used by previous research on Nigeria. The Sample is also split into significant sub samples based on the financial crisis, Global oil increase and Global warming increase period. Furthermore, when it comes to studies on oil price volatility, financial markets and economic activities, focus has been on oil importing and developed economies with the primary direction of their study on stock market returns, output, gross domestic product and industrial production.

However, the primary direction of this study is to examine the long run and short run dynamics of oil price volatility, global financial crisis and macroeconomic leverage in an emerging market economy. According, to the International Finance Corporation of the World Bank Group they suggest that Nigeria is one of their fastest-growing client countries, it is already the ninth-largest in their portfolio and largest exposure in Africa, hence a study on the Nigerian investment and finance sector and factors that affect its performance is of great significance, to policymakers in financial planning and control, and to help investors make optimal portfolio allocation decisions.

In addition, a study of Nigeria is also relevant as it creates a background for studying similar issues in countries such as Algeria, Venezuela, Kazakhstan, and Egypt that are also emerging market economies and oil exporting countries. A study of the Nigerian economy can be used also for comparative analysis for other OPEC member countries. Finally, this study will assist Nigeria in planning for economic growth and development, formulating better monetary and fiscal policies to reduce its level of indebtedness and also useful for investors and financial market participants in making investment decisions and further research.

## 1.5 STATEMENT OF RESEARCH QUESTIONS

What is the relationship between the global financial crisis, oil price volatility and macroeconomic financial leverage in Nigeria?

What is the relationship between macroeconomic financial leverage and Financial Soundness Indicators (country specific and external factor variables)?.

What is the direction of influence between macroeconomic financial leverage and Financial Soundness Indicators (country specific and external factor variables)?.

Do Oil price Shocks have asymmetric and persistent effects on oil price volatility in Nigeria?.

## MAIN RESEARCH PROBLEM1.6 HIERACHICAL STRUCTURE OF THE RESEARCH PROBLEM

## MACROECONOMIC FINANCIAL LEVERAGE, GLOBAL FINANCIAL CRISIS AND OIL PRICE VOLATILITY IN NIGERIA

## RESEARCH QUESTIONS

## What is the direction of influence between macroeconomic financial leverage and Financial Soundness Indicators (country specific and external factor variables)?

## Do Oil price Shocks have asymmetric and persistent effects on oil price volatility in Nigeria?

## What is the relationship between macroeconomic financial leverage and Financial Soundness Indicators (country specific and external factors)?

## What is the relationship between the global financial crisis, oil price volatility and macroeconomic financial leverage in Nigeria?

## ADDRESSING THE RESEARCH QUESTIONS

## The Tests for ARCH Effects, and Exponential Generalized Autoregressive Conditional Heteroscedascity EGARCH.

## The Granger Causality Tests;

## Bilateral Causality, Unilateral Causality and Independence of variables test.

## The Cointegration Technique, Vector Error Correction Mechanism and the Exponential GARCH Model.

## Engle Granger, Durbin Watson, and Johansen Cointegration Technique and The Vector Error Correction Model.

## 1.7 STATEMENT OF RESEARCH HYPOTHESIS

This research carries out four major empirical analysis, the Cointegration Technique, the Vector Error Correction Model (VECM), the Granger Causality Tests and the Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH), in order to achieve the objectives of the study. The research proposes to test the following 43 hypotheses:

## H1: There is a long run equilibrium relationship between Nigerian Government Oil revenue and macroeconomic financial leverage in Nigeria.

## H2: There is a long run equilibrium relationship between Nigerian Financial Deepening Index and macroeconomic financial leverage in Nigeria.

## H3: There is a long run equilibrium relationship between Nigerian Capital Expenditure and macroeconomic financial leverage in Nigeria.

## H4: There is a long run equilibrium relationship between Real exchange rate (NGN/US) [11] and macroeconomic financial leverage in Nigeria.

## H5: There is a long run equilibrium relationship between Energy Consumption Index and macroeconomic financial leverage in Nigeria.

## H6: There is a long run equilibrium relationship between Mineral Production Index and macroeconomic financial leverage in Nigeria.

## H7: There is a long run equilibrium relationship between Consumer Price Index and macroeconomic financial leverage in Nigeria.

## H8: There is a long run equilibrium relationship between Nigerian Industrial Production and macroeconomic financial leverage in Nigeria.

## H9: There is a long run equilibrium relationship between the Nigerian Financial Soundness Indicators and macroeconomic financial leverage in Nigeria.

## H10: There is a long run equilibrium relationship between the United States Gross Domestic Product and macroeconomic financial leverage in Nigeria.

## H11: There is a long run equilibrium relationship between the United States Foreign reserve assets and macroeconomic financial leverage in Nigeria.

## H12: There is a long run equilibrium relationship between United States Consumer Price Index and macroeconomic financial leverage in Nigeria.

## H13: There is a long run equilibrium relationship between United States Imports of goods and services and macroeconomic financial leverage in Nigeria.

## H14: There is a long run equilibrium relationship between United States Industrial production and macroeconomic financial leverage in Nigeria.

## H15: There is a long run equilibrium relationship between United States Government Consumption and Investment and macroeconomic financial leverage in Nigeria.

## H16: There is a long run equilibrium relationship between United States Government Capital Expenditure and macroeconomic financial leverage in Nigeria.

## H17: There is a long run equilibrium relationship between United States Output per hour and macroeconomic financial leverage in Nigeria.

## H18: There is a long run equilibrium relationship between the United States Financial Soundness Indicators and macroeconomic financial leverage in Nigeria.

## H19: The macroeconomic financial leverage and country specific variables can be expressed as a an error correction mechanism.

## H20: The macroeconomic financial leverage and external factor variables can be expressed as a an error correction mechanism.

## H21: Nigerian Government Oil revenue granger causes macroeconomic financial leverage in Nigeria

## H22: Nigerian Financial deepening granger causes macroeconomic financial leverage in Nigeria.

## H23: Nigerian Industrial production granger causes macroeconomic financial leverage in Nigeria.

## H24: Nigerian Energy Consumption granger causes macroeconomic financial leverage in Nigeria.

## H25: Nigerian Principal Mineral production granger causes macroeconomic financial leverage in Nigeria.

## H26: Nigerian Consumer Price Index granger causes macroeconomic financial leverage in Nigeria.

## H27: Government capital expenditure granger causes macroeconomic financial leverage in Nigeria.

## H28: The NGN/US real exchange rate granger causes macroeconomic financial leverage in Nigeria.

## H29: The Nigerian Financial Soundness Indicators impacts macroeconomic financial leverage in Nigeria.

## H30: The United States Gross domestic product granger causes macroeconomic financial leverage in Nigeria.

## H31: The Unites States Foreign reserve assets granger causes macroeconomic financial leverage in Nigeria.

## H32: The United States Consumer price index granger causes macroeconomic financial leverage in Nigeria.

## H33: The United States Government capital expenditure granger causes macroeconomic financial leverage in Nigeria.

## H34: The United States Government Consumption and Investment granger causes macroeconomic financial leverage in Nigeria.

## H35: The United States Government Imports of goods and services granger causes macroeconomic financial leverage in Nigeria.

## H36: The United States Industrial Production granger causes macroeconomic financial leverage in Nigeria.

## H37: The United States Output per Hour granger causes macroeconomic financial leverage in Nigeria.

## H38: The United States Financial Soundness Indicators impacts macroeconomic financial leverage in Nigeria.

## H39: There is evidence of Autoregressive Conditional Heteroscedasticity (ARCH) effects in Oil price returns in Nigeria.

## H40: There is evidence of volatility Clustering in Nigerian Bonny Light Oil Price.

## H41: There is evidence of asymmetric effects of shocks to oil price volatility in Nigeria.

## H42: There is evidence of persistency effects of shocks to oil price volatility in Nigeria.

## H43: The global financial crisis affects the Nigerian economy through the volatility of oil price in Nigeria.

## 1.8 SCOPE AND LIMITATIONS OF THE STUDY

This study focuses on the relationship between oil price volatility, financial crisis and macroeconomic financial leverage in an emerging market economy, employing country-specific and external factor Financial Soundness Indicators (FSIs) and daily oil prices . The Study covers the period of 1970 to 2010. The period is selected because it provides the period in which Nigeria began to experience vicious fluctuations in the price of oil and its revenue, and financial liberalization, it also covers the period of major global financial crisis, oil increase and global warming increase in the world. The major limitation of this study is resource and time constraint.

## RESEARCH METHODOLOGY AND RESULTS

This study for the purpose of achieving its research objectives and questions employs the following methodologies. Following, the Box Jenkins Approach of Times Series Regression analysis, the descriptive and serial correlation tests are obtained. The tests for stationarity, is employed using the Augmented Dickey Fuller (ADF), the Philips Perron (PP), and Kwiatkowski Phillips Schmidt-Shin (KPSS), tests. The tests for Cointegration is carried out employing the Engle Granger (EG), the Durbin Watson (DW) and the Johansen Cointegration (JC) technique. Granger Causality tests is also employed to investigate the direction of influence among the variables. The Vector Error Correction Methodology (VECM), which is a restricted Vector Autoregressive (VAR) system is employed to investigate the short run dynamics of the models. To model oil price volatility, the Exponential Generalized Autoregressive Conditional Heteroscedascity (EGARCH), is employed to investigate the persistency and asymmetric of effects to volatility.

This study provides evidence of a long run equilibrium relationship between Macroeconomic financial leverage and Financial Soundness indicators (Country specific and external factors) it also provides evidence that the discrepancy between the actual or equilibrium value of Macroeconomic Financial Leverage is eliminated or corrected quarterly. This study provides evidence that the global financial crisis impacts the Nigerian economy, through international trade resulting in volatility of oil price, it also impacts Macroeconomic Financial Leverage through Government Oil Revenue. This study also provides evidence of unilateral and bilateral causality between Macroeconomic Financial Leverage and the Financial Soundness indicators (Country specific and external factors). This study also provides evidence of asymmetric and persistency of shocks to oil price volatility increasing from the pre-global financial crisis and global oil increase period to the global financial crisis and global warming increases period.

## 1.10 ORGANIZATION OF THE STUDY

The remainder of this research work is organized as follows: In Chapter 2, I review the literature which is related to oil prices, global financial crisis, financial markets and economic activities. Chapter 3, provides the methodology of the research work. Chapter 4, describes data presentation, results, analysis and interpretation, finally, Chapter 5, presents the Summary, Conclusion and Recommendations and Suggestions for further research.

## 1.11 SOURCES OF DATA

For the purpose of this research study, secondary data was employed from the period 1970 to 2010, such secondary data on the country-specific variables, and external factor variables was obtained from the following:

Central bank of Nigeria [CBN]: Statistical bulletin, 2006,2007,2008.

National Bureau of Statistics: Annual abstract of statistics of Nigeria, 2008-2009.

DataStream, Thomson Reuters Learning.

## CHAPTER TWO

## LITERATURE REVIEW

## 2.1 INTRODUCTION

In the previous chapter, the main focus of this study was extensively discussed, in terms of the statement of research questions, research objectives, hypothesises, the hierarchical structure of the main research problem and the significance of this study. In this chapter, some basic concepts are first reviewed, with an overview of the Nigerian oil industry. Previous literature on Financial Markets, Oil Price Volatility and the Economy is reviewed in two main parts, which are; (A) Theortitical Studies and (B) Empirical Studies. The Empirical studies are divided into three sub sections; focusing on i. the US and other Developed Countries ii. Emerging Markets, Developing Economies and Other Countries iii. Oil Exporting and Oil importing Countries. The Asymmetric effects of oil price changes and the Implications of the Theoretical and Empirical Studies on this study is also analysed.

The final section of this chapter focuses on the Motivation and Contribution of this study to existing literature, this section is also divided into various sub sections examining the following; i. Nigeria as an Emerging Market economy, ii. Oil Price Volatility, iii. Macroeconomic Financial Leverage, iv. The impact of the Global Financial Crisis on the Nigerian Economy, v. Country- Specific and External Factors Effects vi. Dynamics of Macroeconomic Financial Leverage and the Global Financial Crisis. vii. The Asymmetric and Persistency Effects of Oil Price Volatility.

The main shocks that affected the World Economy after World War II were oil price shocks. Therefore, particular consideration has been given to oil price shocks and it affects on financial markets including the economy as a whole. The recent Gulf of Mexico Oil spill also demonstrates that oil is an important factor in worldwide financial and economic stability. There are a number of studies on the relationship between financial markets, oil price volatility and economic activity, more particularly in countries characterized as major exporters or importers of oil, and countries whose main economy is dependent on oil revenue, its changes and effects. The existence of a negative relationship between oil prices and macroeconomic activity became widely accepted since Hamilton’s 1983 work demonstrating that oil price increases reduced US Output Growth between 1948 to 1980. Hamilton’s work has been confirmed and extended by other researchers such as Ferderer (1996) and Uri (1996).

## 2.2 REVIEW OF IMPORTANT CONCEPTS

## 2.2.1 FINANCIAL LEVERAGE, FINANCIAL CRISIS AND OIL PRICE VOLATILITY.

Financial Leverage is generally defined as the extent to which firms utilize debt in financing decisions. It is a very important concept in the whole of finance. According to Clark (2002) [12] in corporate finance, financial leverage plays a major role in determining financial risk, emphasizing that it reflects interest expense causing variability in net income more than that of operating income as a result of operating risk. Considering the macoreconomy, financial leverage also reflects in interest expense causing variability in Gross Domestic Product or Gross National Product, as such where macroeconomic financial risk is concerned the same effect is present. Over several decades, the world has experienced various degrees of financial and economic crisis. The most recent crisis that started in the United States in 2007, has been argued by researchers, economists and politicians to be worse the world has ever experienced since the great depression in 1930s.

A general definition of Financial Crisis is a situation when demand unexpectedly rises relative to money supply, it is characterized by banking and debt crisis, also currency crisis. It is also an important aspect in finance. According to Lee (1995) financial crisis means a recession, characterized by liquidation of the stock market, strain in the banking system, liquidation of bank loans, declining prices, and bankruptcies. Jones (2009) provides a more concise definition, of financial crisis as a sharp ultra-cyclical decline of all or most of a group of financial indicators such as short term interest rates, real and financial assets (stocks, real estate), prices, commercial insolvencies and failures of financial institutions. Accordingly, Mishkin (1999), argued that financial markets perform the crucial function of channelling funds to those individuals or firms that have productive investment opportunities, therefore financial crisis is characterized as a situation where the financial system does not perform this role appropriately, then the economy cannot operate efficiently and economic growth is hampered.

A concept that is of importance in the whole of finance is volatility [13] . Financial and economic time series data such as stock prices, oil prices, exchange rates often exhibit the phenomenon of “volatility clustering” [14] . It signifies that extended periods in which their prices show wide swings are followed by periods of relative calm. Therefore, individuals, firms, government should know how volatility transmissions occur across markets over time, and how they affect their investment and financing decisions. It simply implies that price changes over time, however trading requires volatility, because there will be no need to hedge, and speculate if prices do not change over time. The volatility in the price of oil increases uncertainty in investment decisions, the basic theoretical explanation for oil price volatility is insufficient supply faced with growing demand. Where the sources of oil price volatility are concerned, economists and politicians [15] tend to blame speculators for driving up prices and the Organization of Petroleum Exporting Countries, (OPEC) as a dominant authority in the oil market.

However, empirical research by W.Yang, Hwang, Huang (2002) provide evidence of other factors that play an important role in oil price fluctuations. They argued that the factors that affect the volatility of the United States Oil market are the unstable demand structure, related elasticity of demand and the market structure of OPEC. They argue that most of the post-war recessions were preceded by oil price shocks (1974, 1980, 1990, and 1998), hence volatility of crude oil price creates uncertainty and an unstable economy for both oil exporting and importing countries.

## 2.3 OVERVIEW OF THE NIGERIAN OIL INDUSTRY

Oil was discovered by Shell-BP in Nigeria at Oloibiri in the Niger Delta in the year 1956. This discovery led to the development of the oil industry in 1961. With the quantity of oil been produced increasing, Nigeria attained the status of a major oil producer, ranking 7th largest oil exporting country in the world and the largest oil producing country in the whole of Africa. In the year 1971, Nigeria joined the Organization of the Petroleum Exporting Countries which was created at the Baghdad Conference in Iraq in September 1960. The purpose of the OPEC, as with any cartel is to limit supplies in the hope of keeping prices high to increase the oil revenue of their member countries, this is accounted for why it is often blamed for volatility and increases in the price of oil.

Nigeria holds the largest natural gas reserves in Africa but has limited infrastructure in place to develop the energy sector. The Oil industry plays a dominant role in the Nigerian economy and is regulated by the Ministry of Petroleum Resources. It accounts for about 90% of her gross earnings. However, this dominant role has resulted in the neglect of other relevant sectors in the economy especially agriculture, the traditional mainstay of the economy. The upstream oil industry is a very important sector in the country’s economy providing over 90% of its total exports [16] . There are five major basins in which oil is produced in Nigeria, they are the Niger Delta, Anambra, Benue, Chad and Benin. The major exports crude are BonnyLight [17] and Forcados. Pipeline vandalism and militant attacks [18] are common in these basins especially the Niger Delta. Also common with the basins is oil theft commonly referred to as “oil bunkering” which leads to severe pipeline damage, causing loss of production, pollution and forcing companies to shut down production.

## 2.4 THEORITICAL LITERATURES

This section focuses on the theoretical literatures on the relationship between financial markets, oil price volatility and the economy.

Finance is generally seen as a branch of economics, that focuses on the management of real and financial assets, in particular Modern Financial Theory is founded on three essential assumptions: Markets are highly efficient, Investors are rational, and Investors exploit potential arbitrage opportunities Dimson and Mussavian (1999). Oil is the lifeblood of Modern Economies Sardosky and Basher (2006). Theoretical studies such as Hamilton (1983:1988), Sardsoky and Basher (2006) have provided evidence that changes in the price of oil and its volatility have significant effect on financial markets and economies of the world. Therefore, investors or financial market participants need to know how shocks and its volatility are transmitted across markets over time Malik and Ewing (2009).

Substantial increases in the price of oil are generally acknowledged to have significant effects on both economic activity and macroeconomic policy. Several Economists [19] have offered a number of theoretical explanations to account for the inverse relationship between oil price changes and level of economic activity, they argue that oil price shocks are indicative of increased scarcity of energy.

According to Economic Neo Classical Growth theories, oil price volatility has various implications on macroeconomic activities through both demand and supply channels, known as transmission mechanisms. Aside from supply and demand channels, interest rates, wages, employment, inflation, and foreign exchange markets are other mechanisms through which oil price changes affect the economy.

Balke et. al., (2002), concludes that interest rates are an important mechanism through which oil prices affect economic output. Mussa (2000) also argues that by affecting economic activity, corporate earnings, inflation and monetary policy, an increase in oil price has implications for asset prices and financial markets. Askari (2008) argued that oil price dynamics for the period 2002-2006, was characterized by high volatility, high intensity jumps, strong upward drifts, pressure on oil prices resulting from rigid crude oil supply and expanding world demand for crude oil.

Fluctuations in oil prices are caused by supply and demand imbalances arising from events like changes in political regimes, economic and financial crises, organization or termination of trade agreements. According to Sardosky and Basher (2006), the demand side effects, reflects in consumption and investment, arguing that higher oil prices take the form of inflation tax on consumers and producers by reducing the amount of disposable income that consumers have left to spend on other goods and services, also raising costs of non-oil producing companies, reducing profits and dividends. The supply side effects is argued on the basis of oil as an significant input in production, oil price increases can lead to decreased productivity of other inputs which consequently leads to lower output.

The earliest studies on the effects of oil prices on financial markets performance and the economy is attributed to Hamilton (1983;1988), evidence was found that energy price shocks reduce aggregate employment by inducing workers in adversely affected sectors to remain unemployed, Mork (1989) also discovered the same findings which is consistent with economic growth theory. An imperfect competition model accounts for the effects of oil price shocks on declining real wages and output, this was discovered by Rotemberg and Woodford (1996). Chen, Roll and Ross (1986), are also among the first researchers to systemically investigate the impact of macroeconomic variables on stock price returns. They explain that a long run equilibrium relationship exists between financial markets and the macoreconomy. Moreover, their findings show that interest rates, inflation rates, bond yield spreads and industrial production are affected by oil price shocks.

However it is commonly argued that oil price shocks have diverse effects on various sectors and industries. Empirical evidence provided by Lee and Ni (2002), show that all sectors are not equally affected by oil shocks, in consistent with their findings, Faff and Nandha [20] (2007) also provide evidence that oil price rises have a negative impact on equity returns for all sectors except mining, oil and gas industries.

## 2.5 EMPIRICAL LITERATURES

This section of the chapter focuses on empirical literatures on the relationship between financial markets, oil price volatility and the economy, studies on the United States and other developed economies are first reviewed, Emerging Markets and other Developing Countries, Oil Exporting and Importing Countries are also reviewed. Asymmetric effects of oil price shocks is also reviewed.

## 2.5.1 US and DEVELOPED ECONOMIES

The United States is commonly referred to as a World Power because of its importance in the economic performance of other countries. Hence, using a Vector Autoregressive System (VAR) Hamilton (1983) found statistically significant correlation between oil price shocks and economic recessions in the US economy. Burbridge and Harrison (1984) also employed a VAR model with monthly data from May 1962 to June 1982 for the US, Japan, Germany, Canada and the UK. According to the results of the impulse response analysis, the impact of oil price shocks on industrial production on UK and US is substantial while in Japan, Germany and Canada it is relatively small.

Another study on the United States was carried out by Uri (1996), the effect of changes in the price of crude oil on agricultural employment was investigated between the period 1947 and 1995. The existence of an empirical relationship between agricultural employment and crude oil price changes was established using the Granger Causality methodology. Cologni and Manera (2007) employed a structural cointegrated VAR model inorder to study the direct effects of oil price shocks on output, prices, money demand, exchange rate and interest rates for G7 [21] Countries, they found evidence of oil price influences on long run equilibrium relationships for all countries except Japan and the United Kingdom.

In order to simultaneously estimate the mean and conditional variance between five different US Sector Indexes and Oil Prices, a different methodology from previous studies was employed by (Malik and Ewing (2009)). They employed a bivariate GARCH [22] and found evidence of a significant transmission of shocks and volatility between oil prices and market sectors. According to them, different financial assets are traded based on different market sector returns, it is therefore important for financial market participants to understand volatility transmission over time and across these series in order to make optimal portfolio allocation decisions.

## 2.5.2 EMERGING MARKETS, DEVELOPING ECONOMIES and Other COUNTRIES

Emerging markets are becoming increasingly important to wealthy investors [23] and financial market participants. In terms of risk profiles they are categorized in the high risk profile on the basis of high risk, high return securities. Hence, studies on emerging markets are of great importance in the finance and investment sector. A multi factor model that allows for both unconditional and conditional risk factors was employed by Sardosky and Basher (2006) to study the impact of oil price risk on a large set of emerging stock markets. According to them, emerging economies tend to be more energy intensive than advanced economies and are therefore more exposed to higher oil prices. Consequently, oil price changes is likely to have a greater impact on profits and stock prices in emerging economies than other economies. They find strong evidence that oil price risk impacts stock returns in emerging markets. The increased flow of portfolio money (stocks, mutual funds) in emerging markets securities means that oil price impacts on emerging stock markets affects both domestic and international investors.

A study on the Australian industry equity returns to an oil price factor, was carried out by Faff and Brailsford (1999), employing an augmented market model with monthly data of over the period 1983 to 1996, they found a positive and significant impact of oil prices on the Oil and Gas, Diversified Resources industries and a negative and significant impact of oil prices on Paper, Packaging and Transportation industries. The result is also consistent with Lee (2002), Faff and Nandha (2007), who found that different sectors react in different ways to oil price shocks.

In order to explain the dynamic relationship among oil prices, real stock prices, interest rates, real economic activity and employment for Greece, Papapetrou (2001) employed a Vector Autoregressive System. The results suggests that oil prices affects real economic activity and employment, also oil prices are important in explaining stock price movements; however contrary to previous evidence stock returns do not lead to changes in real activity and employment in Greece.

A comprehensive study was carried out on many European Countries and some Asian countries by Cunado and Gracia (2005), employing Cointegration and Granger Causality tests for the period 1975 to 2002 [24] . They analysed the relationship between oil prices and macroeconomic variables such as inflation and production growth with economic activity. The world oil price and national oil price, Consumer Price Index, are employed, economic activity is estimated with industrial production data. For oil price changes three specifications are used the Real Oil price changes, Net Oil Price Increases and the Scaled Oil Price Increases. All the variables contain a unit root [25] so first differences are taken, however there are no Cointegration among the variables, which means there is evidence of no long run relationship between oil price and the economic and financial variables, for the selected countries which is contrary to previous research evidence.

A Vector Error Correction Model (VECM), was employed by Rano Aliyu (2009), in studying the impact of oil price shock and real exchange volatility on real economic growth in Nigeria, evidence was found that real exchange rate and oil price changes exerts a positive influence on economic growth. Sari and Soytas (2006), also found contrary results, as oil price shocks do not appear to have significant impact on real stock returns on Istanbul Stock Exchange they investigated the impact of oil price shocks on the real stock returns, employment, Gross Domestic Product GDP and inflation in Turkey employing the VAR methodology.

Similarly, Rafiq, Salim and Bloch (2008) and Rong-Gang-Cong, Yi-,Ming Wei, Jian-Lin Jiao, Ying Fan (2008), employ a Vector Autoregressive model to examine the impact of oil price volatility on key macroeconomic and financial indicators of Thailand and China respectively. The results of the Granger Causality test, impulse response functions and variance decompositions show that oil price volatility has significant impact on macroeconomic indicators such as unemployment and investment in Thailand over the period from 1993Q1 to 2006Q4. Using UK Brent crude oil price as a representative of world real oil price, Rong-Gang-Cong et. al., (2008), found that oil price shocks do not show statistically significant impact on the real stock returns of most Chinese stock markets indices, except for manufacturing index and some oil companies for the period 1996 to 2007 also providing evidence that oil price shocks does not equally affect all sectors.

## 2.5.3 OIL IMPORTING AND EXPORTING COUNTRIES

According to the Central Intelligence Agency, of the United States of America [26] , the ranking of top ten World Oil Importers, are the United States, European Union, Japan, China, South Korea, India, Germany, Netherlands, France, Italy while the ranking of top ten World Oil Exporters are Saudi Arabia, Russia, United Arab Emirates, Iran, Kuwait, Norway, Nigeria, Angola, Venezuela, and Algeria.

A number of researchers have focused more on Oil importing countries, especially the United States and European Union, and concluded that oil price increases have a negative impact on economic activities, establishing a long run equilibrium relationship. Among these researchers include Hamilton (1983), Burbidge and Harrison (1984), Uri (1996), Ferderer (1996), W.Yang, Hwang, Huang (2002), Hui Guo and Kevin L. Kliesen(2005), Cologni and Manera (2007), Malik and Ewing (2009).

A comprehensive study was carried out by Hammoudeh and Elesia (2004), on the relationship between oil price and stock prices for five members of the Gulf Cooperation Council (GCC),- Bah ram, Kuwait, Oman, Saudi Arabia and United Arab Emirates, using daily data they find that only Saudi Arabia, had a bi-directional relationship between oil prices and stock prices. Ayadi (2000), Olomola and Adejumo (2006), examined oil price shock on macroeconomic activities in Nigeria, they examine gross domestic product, output, inflation, real exchange rate and money supply, the Vector Autoregressive Methodology was employed to analyse the data. Their findings were contrary to previous empirical findings in other countries, oil price shocks does not affect output and inflation in Nigeria for the period under study 1970-2003, however it significantly influence real exchange rates, implicating a high real oil price may give rise to wealth effect that appreciates the real exchange rate.

In order to examine the long run and short run dynamics of oil price and monetary shocks for the Russian Economy, Ito (2008), employed a VECM-Vector Error Correction approach, covering the period between 1997Q1 and 2007Q4 and found that increases in oil prices positively impacts Gross Domestic Product. Similarly, using a Vector Error Correction Model (VECM), Oskooe (2010) investigates the relationship between stock market performance and economic growth in Iran, finding evidence that real economic activity including that of the oil sector affects the movement of stock prices.

## 2.6 ASYMMETRIC EFFECTS OF OIL PRICE SHOCKS

The literature on oil price shocks suggests the importance of information contained in oil price shocks for financial markets may be different for both developed and emerging economies, also for various sectors of the economy. Asymmetry in effect means that oil price increases have a apparent negative impact on economic activity while oil price decreases don’t affect economic activity significantly. Tatom (1988) provides evidence that monetary policy has responded asymmetrically to oil price shocks. Mork (1989) also provides evidence of oil price decreases not having a statistically significant impact on the US economy. Thus, an asymmetric effect is apparent because real effects of oil price decreases are different from those of oil price increases. Asymmetric effects of oil price changes was also confirmed Mork, Olsen and Mysen (1994), study for the Organization for Economic Co-operation and Development (OECD) [27] Countries. A study of asymmetric effects was also carried out by Ferderer (1996) and Hooker (1996) they provided evidence that monetary policy responds to oil price increases and not to oil price decreases, also response of short term interest rates to oil price increases and decreases is asymmetric.

## 2.7 IMPLICATION OF THE THEORITICAL AND THE EMPIRICAL LITERATURE FOR THIS STUDY.

Most of the previous empirical studies have focused on oil importing countries and developed economies, United States, in particular. Not many studies exist on the effects of oil prices, its volatility for major oil exporting countries, and emerging market economies. In addition, majority of the studies on financial markets, oil price volatility and the economy have focused primarily on productivity, inflation, stock market returns, gross domestic product, and real exchange rate, however, evidence does not exist for macroeconomic financial leverage, and oil price volatility in Nigeria. Rafiq et. al., (2008), argued that studies analysing oil price volatility are needed for emerging market economies. Narayan et. al.,(2007), scarce literature exists on oil price volatility. This study intends to fill this gap.

## 2.8 MOTIVATION AND CONTRIBUTION TO LITERATURE.

This section focuses on the motivation and contribution of this study to existing literature in terms of the primary direction of the study, methodology of the research and the data set to be employed.

## 2.8.1 NIGERIA AS AN EMERGING MARKET ECONOMY

Nigeria as a country has certain features that makes it important and interesting to be studied. It is categorized as an Emerging Market Economy in the Emerging Market Data Base (EMDB) [28] , and heavily dependent on Oil Revenue, it is also Africa’s leading producer of oil, and is the 7th Largest Exporter of Oil in the world, it is a member of the Organization of Petroleum Exporting Countries [29] , with United States as the major importer of Nigerian Oil, any problem that emanates from the United States economy, affects the Nigerian economy, it is among the major African borrowers of finance institutions such as the World Bank, and African Development Bank, it is also the most populous country in Africa, with over 148million people [30] .

## 2.8.2 Oil Price Volatility in Nigeria.

The effect of oil price volatility on the Nigerian economy has only been studied by a few researchers and the primary direction of their study was on its impact on the Gross Domestic product, Oil revenue, Real exchange rate, and Inflation. The economic a prior expectation is that fluctuations in oil do significantly affect aggregate economic activities, because of the over dependency of oil revenue by the economy. This research contributes to previous study and literature by focusing on macroeconomic financial leverage and oil price volatility in Nigeria.

Furthermore, this research also contributes to literature by studying the asymmetric and persistency of shocks to oil price volatility, employing the Exponential Generalized Autoregressive Conditional Heteroscedascity (EGARCH) technique. The rationale behind this investigation is that previous studies by Ayadi et al (2000), Rano Aliyu (2009) have provided empirical evidence on a positive relationship with oil revenue, output, gross domestic product and oil price increases. This suggests that oil price increases exert positive influence on oil revenue and gross domestic product for the period 1986 to 2007(Rano Aliyu 2009) and 1975-1992(Ayadi 2000). Olomola and Adejumo (2006) also provide empirical evidence that oil price shocks affect real exchange rates substantially, and it does not affect inflation in Nigeria for the period 1970 to 2003.

However, empirical evidence has not been provided on the asymmetric, and persistency of shocks to oil price volatility, the effect of oil price volatility on macroeconomic financial leverage in Nigeria. An important issue facing policy makers for several decades is the volatility of oil price , its causes and effects on the Nigerian economy. Investors also need to know how the oil price volatility can affect their investments. Thus, a study on oil price volatility and macroeconomic financial leverage in Nigeria is relevant for both policymakers and investors.

## 2.8.3 Macroeconomic Financial Leverage in Nigeria.

It is widely known that most emerging markets of the world rely substantially on external funding in financing its development projects [31] . This external funding usually takes the form of external loans as an element of macroeconomic financial leverage. Macroeconomic Financial leverage in Nigeria is mainly categorized into Domestic Debt, External Debt [32] and Government Treasury Bills. The Nigerian Government issues more of treasury bills as debt market instruments so it goes a long way to determine their financing decisions.

According to the Nigerian Debt Management office, external debt as at March 30th 2010 owed to the World Bank, African Development Bank was USD4.3billion which grew from USD3.947billiion year end 2009, domestic debt as at March 30th 2010 was USD3.5billion which grew from USD3.2billion year end 2009. Adofu, Audu and Abula (2010), provided empirical evidence of a negative relationship between economic growth and domestic debt in Nigeria. Adegbite, Ayadi et. al., (2006), also provide evidence that external debt affects economic growth. An important issue facing policy makers for several decades is the year on year increase in macroeconomic financial leverage, its causes, sources and effects on the Nigerian economy.

In developed economies, consider the United States with organized regulatory systems the main source of financing for Government spending is income tax, supplemented by borrowing from the public. However in an emerging market economy like Nigeria, the oil revenue not income tax is the main source of Government spending, supplemented by domestic and external borrowings. Consequently, it is at risk to fluctuations in the international price of oil, and when there is too much instability in the price of oil the government cannot project its revenue accurately, and investment spending cannot be well planned in advance, this is a major concern of oil exporting countries that depend heavily on oil revenue for their public projects. Thus, majority of emerging markets more importantly oil exporting countries are faced with obtaining external, and domestic funds to supplement oil revenue.

As a prominent member of the Organization of Petroleum Exporting Countries OPEC, the Nigerian Government has to honour the quota system imposed by the organization and because crude oil is a publicly traded community, its price is determined in the commodity markets through the interaction between worldwide demand and supply. Nigeria has no control over the price or the quantity making it very difficult for its government to project annual revenue, as such financial leverage by the Government is to supplement oil revenue. Macroeconomic financial leverage, in this study is categorized as Government Domestic Debt, Government External Debt, and Interest rate on Treasury Bills. Proper examination and analysis of the different channels or mechanism of transmission between the external shocks caused by oil price changes and domestic economic performance becomes a very significant subject matter for economic policy formulation and implementation.

## 2.8.4 The impact of the Global Financial Crisis on the Nigerian Economy.

A further contribution of this research, is to study the impact of the global financial crisis on macroeconomic financial leverage and oil price volatility in Nigeria. A widely accepted argument is the global financial crisis affects the Nigerian economy through foreign direct investment, leading to the weakening of corporate and project finance, (Adamu (2007)). The effect of the crisis on the economy varies from country to country. The wide spread effect for emerging market economies is characterized majorly has weakening export revenues, pressures on current accounts and balance of payments, lower growth translating into higher poverty, weaker health systems, more crime and difficulties in meeting Millennium Development Goals.

A major effect on the Nigerian economy is the volatility in the price of crude oil, and its effect on oil export revenue, and inflow of capital into Nigeria cannot be overemphasised. Eva (2007), argued that oil prices are more volatile than prices of other commodities, using monthly data from the period January 1945 to August 2005 on oil prices, concluded that the dynamics of oil prices are more important for oil producing and exporting countries.

This study examines the effect of global financial crisis on oil price volatility by examining the persistency and asymmetric of shocks to volatility during the pre-crisis and the global financial crisis period. The sample period is also selected in reference to the 2000s Global oil increase and 2006s Global warming increase period. Inorder, to examine the effect of the global financial crisis on macroeconomic financial leverage, the short run and long run dynamics of Government Domestic Debt, Government External Debt, Interest rate on Treasury Bills and the Financial Soundness Indicators as developed by the International Monetary Fund, World Bank and European Central Bank is investigated. The Financial Soundness Indicators employed in this study are based on Country- Specific- Nigeria and External Factors-United States factors.

## 2.8.5 The Country-Specific and External Factor Effects.

The United States as a major importer of the Nigerian oil, implies that the shape of the United States economy affects overall economic performance. The Global financial crisis began in 2007, with the subprime mortgage market in the United States. There are multiple channels by which a slowdown in the US economy can be transmitted to the Nigerian economy. Therefore, it is important to investigate if the status of the US economy affects the Nigerian economy. Thus, this study also contributes to research, by examining the relationship between external factor (United States), variables and macroeconomic financial leverage in Nigeria.

Accordingly, this research examines the relationship between macroeconomic financial leverage, and financial soundness indicators for country-specific and external factors effects. The financial soundness indicators are Index of Industrial (manufacturing, mining, electricity) production and Mineral (Petroleum, Gas, Cassiterite, Columbite, Coal, Limestone) production, Energy Consumption, Output per hour to measure productivity, Government Oil Revenue to measure profitability, Government Capital Expenditure, Import of goods and services, Gross Domestic Product, Foreign Reserve Assets, Consumer price index to measure domestic price level, Real Exchange rate which is the average of official exchange rate of the Naira vis a vis the United States Dollar, and Financial Deepening [33] index is to measure growth opportunities of the economy.

A number of theoretical and empirical studies have provided evidence that profitability, growth, tangibility and volatility affects financial leverage, Friend and Lang (1988), Titman and Wessels (1988), for United States firms, Kester (1986) for United States and Japan, Rajan and Zingales (1995), Wald (1999) for developed countries, Wiwattankantang (1999) for developing countries. The major theories that explain firm financing are the Pecking order developed by Myers (1984) and Trade off theory. The Pecking order theory, based on the premise that firms follow an hierarchical form of financing, whereby the cheapest sources of finance are first utilized, internal funds followed by debt and then equity. It can be argued that with Nigeria as a major oil dependent country, a financial deficit in oil revenue leads to debt by the Government in form of external and domestic debt, and government securities.

In the research work by Nzotta and Okereke (2009) they employed financial deepening as a measure of Government growth opportunities, which is the extent to which financial institutions mobilize savings for investment and development purposes. They investigated its effect as a necessary condition for economic growth. A high financial deepening index indicates high growth opportunities for the economy, as such low levels of debt. This research also contributes to existing literature by examining the relationship between financial deepening index and Macroeconomic financial leverage.

A number of researchers have argued that oil is a major asset used as collateral by the Nigerian government in financial leverage Olomola and Adejumo (2006)). Therefore Nigeria being the largest producer of oil in Africa suggests that it has a high fraction of tangible assets in terms of oil, consequently there is a huge incentive to be in debt. It can be argued that this accounts for the year on year increase in debt as reported by the Debt Management office of the Nigerian government.

Additionally, to the relevance of this study to economic and policy formulation, it is also vital for investors who spread their portfolios across emerging markets securities to understand the economic and financial characteristics of these emerging markets . In line with the theory of asset pricing, high expected returns should be associated with large exposure to risk factors, consequently emerging markets are seen as high risk , high return markets.

A remarkable study was carried out by Harvey (1995) employing data on more than 800 equities in six Latin American Markets (Argentina, Brazil, Chile, Colombia, Mexico and Venezuela), eight Asian Markets (India, Indonesia, Korea, Malaysia, Pakistan, Philippines, Taiwan and Thailand), three European Markets (Greece, Portugal and Turkey), one middle east market (Jordan), two African markets (Nigeria and Zimbabwe) [34] . He found that the addition of emerging market assets significantly enhances portfolio opportunities. Sardosky and Basher (2006) and Eryiğit (2009) found strong evidence that oil price risk impacts stock price returns in emerging markets. Therefore high volatility in oil price suggests uncertainty about the future return on investment, which may affect the present value of dividend payments.

## 2.8.6 Dynamics of Macroeconomic financial leverage, and Global financial crisis in Nigeria.

Previous researches by Ayadi (2000), Olomola and Adejumo(2006), Omisakin, Adeniyi and Omojolaibi (2009), Rano Aliyu (2009) on oil price volatility and financial, macroeconomic variables have all employed similar research methodologies. Ayadi(2000), employed the Vector Autoregressive Methodology (VAR), in studying the effect of oil price changes on the Nigerian economy, using the variables, gross domestic product, inflation and real exchange rate. Olomola and Adejumo (2006), employed Johansen Cointegration tests and also the Vector Autoregressive Methodology, in studying the oil price shock and macroeconomic activities in Nigeria, using the variables real gross domestic product [35] , Consumer price index, real exchange rate [36] , wholesale price index and real oil price [37] .

Omisakin et. al., (2009) also employed the Johansen Cointegration and the Vector Error Correction Model using the variables; gross domestic product, energy price [38] , government expenditure, oil revenue, money supply and consumer price index. Aliyu (2009), also employed a similar methodology, however also carried out Granger causality tests in studying the impact of oil price shock and real exchange volatility on real economic growth in Nigeria.

The contribution of this study in terms of the research methodology is also very significant, the tests for stationarity are also carried out using the Augmented Dickey Fuller and Philips Perron tests as with previous research, the Kwiatkowski Phillips Schmidt-Shin test is also employed to test the robustness of the ADF and PP results. The Durbin-Watson approach [39] , the Engle-Granger approach and Johansen approach is included in the tests for cointegration, for the purpose of testing the robustness of the results. In consistent with previous studies to examine the short run dynamics of financial leverage, financial crisis and oil price volatility, the Vector Error Correction methodology is employed as it reconciles short run and long run behaviour. Also consistent with previous research the Granger Causality tests, is employed to examine the direction of influence between the variables in the study.

## 2.8.7 Asymmetry and Persistency of Shocks to Oil Price Volatility in Nigeria.

This study further contributes and extends previous literature by testing for volatility in the BonnyLight Nigerian oil price, by employing the Exponential Generalized Autoregressive Conditional Heteroscedascity (EGARCH). The ARCH model was originally developed by Engle (1982) . The GARCH model was originally proposed by Bollerslev (1986). The EGARCH model was proposed by Nelson (1991). According to Narayan and Narayan (2007), scarce literature exists on oil price volatility. Therefore this study adds to existing literature on the volatility of crude oil price.

In order to model the volatility of real oil BonnyLight price, daily data that has been adjusted is employed from the period 30th of November, 2000 to 7th of July, 2010 is employed. Following the methodology of Narayan et. al., (2007), who argued for carrying out the first study on oil price volatility employing daily data from the period 13th of September 1991 to 15th of September 2006 and finding evidence of persistency and asymmetric effects to volatility in the price of oil using the Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH) methodology.

This Study also argues for carrying out the first study on oil price volatility in Nigeria employing daily data and the Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH) methodology. The sample period is split into two distinctive periods to examine the effect of the pre-crisis, global oil increase, global financial crisis and the global warming increase period on oil price volatility.

## CHAPTER THREE

## RESEARCH METHODOLOGY

## 3.1 INTRODUCTION

In the previous chapter, theoretical and empirical literatures on Financial Markets, Oil Price Volatility and the Economy were reviewed with respect to the various methodologies employed and countries examined, with important concepts also defined and explained. The Motivation and Contribution of this study to theoretical and empirical literature was also extensively discussed in terms of the primary direction of the study, the methodology to be employed, and the data set to be employed.

This chapter comprises of the methodology of this study, in the view of achieving its research objectives. The model specification, test procedures, data and analytical techniques are discussed. The main purpose of this study is to examine the relationship between oil price volatility, the global financial crisis and financial leverage in the Nigerian economy. Therefore the research methodology [40] is directed towards achieving the research objectives of this study. The Descriptive Statistics and Serial Correlation tests, are first carried out.

Following, the Box Jenkins methodology, the time series is tested for Stationarity, employing the Augmented Dickey Fuller (ADF), The Philips Perron (PP) and Kwiatkowski Phillips Schmidt-Shin test (KPSS). Following, the approach of Modelling Long Run Relationships in Finance, The Cointegration technique is employed using the Engle Granger, Durbin Watson and the Johansen Methodology. A main innovation is that various tests are carried out to judge the robustness of the results For the purpose of Modelling Short Run Dynamics, The Vector Error Correction Model (ECM) is then estimated to reconcile long run and short run behaviour. For the purpose of Modelling Causality in Finance, The Granger Causality tests are also employed, to determine the direction of influence of the variables in the study.

For the purpose of Modelling Volatility, examining the asymmetric and persistency of shocks to volatility, the Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH), is employed to model Oil Price Volatility in Nigeria. Another main innovation, is that Sub- samples are employed to judge the robustness of the results. The Sub- samples are based on distinctive periods such as the Pre- Financial Crisis period, the Global Oil Increase period, the Global Warming increase period and the Global Financial Crisis period.

The Stability and Residual tests are also carried out, the Breusch Godfrey tests is employed for Serial Correlation which is common feature with time series data, Another important test is specification error, however when testing for cointegration if a relevant variable has been omitted from the equation, the test should fail to find cointegration among the variables [41] . In addition, an important assumption with the vector autoregressive system, is that all the variables in the system are endogenous. So, there is no need to test for Multicolinearity. All estimations are carried out using Bollerslev-Wooldridge Standard errors to obtain robust estimates and Newey West Heteroscedasticity and Autocorrelated Corrected Standard Errors, is employed to correct for serial correlation, it’s an extension of White’s Heteroscedasticity Consistent Standard Errors

## 3.2 BOX JENKINS METHODOLOGY

One of the important types of data used in empirical analysis is the times series data [42] . Regression models involving time series data are widely used for economic forecasting. However, a very important characteristic of times series data is that they are non-stationary because they grow over time, which implies that the mean, variance, autocovariances are time variant. [43] Regression analysis is commonly defined as the study of the dependence of one variable, usually referred to as the dependent variable [44] , on one or more independent variables or the regressors.

45Times series models are usually a-theoretical, implying that their construction and use is not based upon any underlying theoretical model of the behaviour of a variable. Time series analysis is sometimes referred to as the [46] Box Jenkins modelling by relying on the past behaviour of the variable being modelled. Assuming Y is the variable to be modelled, Y is transformed to ensure that it is stationary implying that the mean of Yt,, its variance and its covariances with other Y values, for example Yt-k does not depend on (t). Based on the Box Jenkins approach, time series data can be made stationary by differencing, this creates a new data series, Y* , which becomes the input for the Box Jenkins analysis.

Y*t = ϕ1Y*t-1 + ϕ2 Y*t-2 +......+ ϕpY*t-p + Ɛt + ɵ1 Ɛt-1 + ɵ2 Ɛt-2 + .....+ ɵq Ɛt-q...Equation (1)

Where the ϕ and ɵ are unknown parameters and the Ɛ are independent and identically distributed normal errors with zero mean. This is a general model commonly called an autoregressive integrated moving average, ARIMA (p,d,q) model, p is the number of lagged values of Y*, representing the order of the autoregressive (AR) dimension of the model, d is the number of times Y is differenced to produce Y*, and q is the number of lagged values of the error term, representing the order of the moving average (MA) dimension of the model.

## 3.2.1 Ordinary Least Squares

A common phenomenon in regression analysis is the Ordinary least squares technique. However, it is the unique statistical properties of this technique that has made it a very powerful and popular. [47] The estimators of this technique are referred to as the least square estimators. The statistical properties of these estimators based on the [48] Gauss- Markov theorem, are the best linear unbiasedness property of an estimator commonly referred to as the BLUE properties. Linearity simply implies the linear function of a random variable, unbiased means that the average or expected value is equal to the true value. An estimator is referred to as efficient if it has minimum variance in the class of all such linear unbiased estimators. In this research the OLS is the method of estimation for the tests for Cointegration among the variables, tests for ARCH effects and the Error Correction Mechanism.

## 3.2.2 Maximum Likelihood

An alternative method to the ordinary least squares is the maximum likelihood, however this technique can be applied to regression models that are non-linear in parameters as opposed to the ordinary least squares. The maximum likelihood estimation involves forming a log likelihood function, usually denoted LLF, and finding the parameters that maximises the LLF. Maximum likelihood is generally a large sample method, under the normality assumption the maximum likelihood or ordinary least squares estimators of the intercept and slope parameters of the regression model are usually identical. For this research the method of the maximum likelihood is employed in the Autoregressive Conditional Heteroscedasticity modelling for testing the hypothesis of the volatility in oil price, because the model is not linear in parameters the method of the Ordinary least squares is inappropriate.

## 3.3 DESCRIPTIVE STATISTICS

The importance of the descriptive statistics is to give a thorough description of the nature of the data set to be employed in this study. It presents the possibility of the existence of outliers and non-normality of the data set. With measures like the skewness, kurtosis, the mean, median and standard deviation, it presents the data adjustments that is required in the study. The Descriptive statistics is the first step in this research methodology to analysis the nature of data to be employed.

## 3.4 SERIAL CORRELATION TESTS

A common problem with time series data is serial correlation [49] . This is defined as “correlation between members of series of observations ordered in time, for the time series analysis. This is a violation of one of the Classical regression model assumption, which is that the disturbance term relating to any observation is not influenced by the disturbance term relating to any other observation. This is represented below;

cov (ui , uj│xi , xj) = E (ui , uj ) = 0 i ≠ j ..................................Equation (2)

However, if there is such a relationship, we have autocorrelation; Symbolically, this is represented below;

E (ui , uj ) ≠ 0 i ≠ j....................................Equation (3)

The main common causes of serial correlation is specification bias, especially the case of an excluded variable and an incorrect functional form. The manipulation of data and incorrect data transformation can also lead to serial correlation. Non-stationarity of data is also a possibility the problem. The consequences of ignoring serial correlation, are; the Ordinary least squares estimators are still unbiased and consistent but inefficient, variances of coefficients are biased and tests are invalid. The Breusch Godfrey test is employed to test for serial correlation in this study.

The Newey- West Standard errors are employed in this study to correct for serial correlation, and obtain standard errors of Ordinary least square estimators that are corrected for serial correlation. The Standard errors are also corrected for Heteroscedasticity, although it is not a common feature with Time series data

## 3.5 TESTS FOR STATIONARITY

The Box- Jenkins approach is only valid if the underlying time series is stationary. The phenomenon of a “random walk” is common with times series data. Therefore, an important assumption based on times series regression is that the underlying times series is stationary. It is widely accepted that a time series is stationary, if the mean, variance and autocovariances are time invariant [50] . A popular test of times series stationarity is the unit root test.

For this research, the Augmented Dickey Fuller (ADF) test, Philips Perron (PP) test and Kwiatkowski Phillips Schmidt-Shin test (KPSS) will be employed to test for unit root. Standard ADF and PP test are not very informative on how to distinguish between a unit root and near unit root case, the KPSS test proposed by Kwiatkowski, et al., (1992), is employed in a complementary way to confirm the results of the ADF and PP tests. The hypothesis for the KPSS are a reverse to the hypothesis of the ADF and PP tests. The regression equation can contain a constant, or a constant and trend. The augmented dickey fuller test protects against the possibility of a autocorrelated error. It is important to also note that the ADF and PP, and KPSS tests are unaffected by Heteroscedasticity [51] . A basic assumption of the Dickey Fuller tests is that the disturbances are independently and identically distributed, Philips Perron (1988) [52] , proposed the Z statistics, with the assumption off IID [53] relaxed.

The equation is specified as:

Xt = ϕ1Xt-1 + ϕ2 Xt-2 +......+ ϕqXt-q + Ɛt ....................................................Equation (4)

∆Xt = δXt-1 + δ1 Xt-1 + δ2∆Xt-2 +..........+δq-1 ∆Xt-q+1 + Ɛt........................ .Equation (5)

The Hypothesis for the ADF and PP [54] is formulated as follows:

H0: δ=0 (there is a unit root or the time series is nonstationary tδ > T).

H1: δ<1 (there is no unit root or the time series is stationary tδ < T).

If the computed tau statistics [55] /ADF/PP statistics exceed the ADF/PP critical tau values [56] , we reject the null hypothesis, which implies that the time series data is stationary. In other words for the times series to be stationary, the tδ must be much negative than the critical values.

## 3.6 COINTEGRATION

A general definition for cointegration is “A vector of variables integrated of order one is cointegrated if a linear combination of the variables are stationary”. A variable is said to be integrated of order d, written as I(d), if it must be differenced d times to induce stationarity, thus a stationary variable is integrated of order zero I(0). It is commonly known that regression of nonstationary series on stationary series can lead to spurious regressions. A statement popularly attributed to Scientist Granger (1962), is that “a test for Cointegration can be thought of as a pre-test to avoid spurious regression situations”.

Cointegration among variables simply means that a long run or equilibrium relationship exists among the variables. It provides a framework for testing and estimating long run or equilibrium relationships among variables. If the time series contains a unit root, differencing might lead to loss of long run information, therefore a test for cointegration is very important, if a cointegrating relationship is found, an ECM- Error correction model is estimated.

## 3.7 ERROR CORRECTION MECHANISM

There could be disequilibrium in the short run, therefore the error term can be treated as equilibrium error. The ECM is a means of reconciling short run and long run behaviour. Granger representation theorem also states that if two variables are cointegrated, the relationship between the two can be expressed as an [57] Error Correction Mechanism (ECM).

For example we consider the following relationship;

yt = β0 + β1xt + β2xt-1 + β3yt-1 + Ɛt.............................................................Equation (6)

where y and x are measured in logarithms, with economic theory suggesting that in the long run y and x will grow at the same rate, so that in equilibrium (y - x) will be a constant, except for the error. For this study the equations are specified as:

## Country- Specific Equations- Nigeria Indicators

NGNGDDt = β0 + β1NGNFDt + β2NGNCPIt-1 + β3NGNREAXt-1 + β4NGNIDPt-1 + β5NGNIMPt-1 + β6NGNIECt-1 + β7NGNCAPEXt-1 + β8NGNGOILRt-1 + Ɛt .......Equation (7)

NGNGEDt = β0 + β1NGNFDt + β2NGNCPIt-1 + β3NGNREAXt-1 + β4NGNIDPt-1 + β5NGNIMPt-1 + β6NGNIECt-1 + β7NGNCAPEXt-1 + β8NGNGOILRt-1 + Ɛt ........Equation (8)

NGNINTBILLt = β0 + β1NGNFDt + β2NGNCPIt-1 + β3NGNREAXt-1 + β4NGNIDPt-1 + β5NGNIMPt-1 + β6NGNIECt-1 + β7NGNCAPEXt-1 + β8NGNGOILRt-1 + Ɛt .........Equation (9)

## External Factors Equations: US indicators

NGNGDDt = β0 + β1USCPIt + β2USGDPt-1 + β3USIMPORTSt-1 + β4USCAPEXt-1 + β5USIDPRODt-1 + β6USFORRESt-1 + β7USGCINVt-1 + Ɛt ..................Equation (10)

NGNGEDt = β0 + β1USCPIt + β2USGDPt-1 + β3USIMPORTSt-1 + β4USCAPEXt-1 + β5USIDPRODt-1 + β6USFORRESt-1 + β7USGCINVt-1 + Ɛt ....................Equation (11)

NGNTBILLt = β0 + β1USCPIt + β2USGDPt-1 + β3USIMPORTSt-1 + β4USCAPEXt-1 + β5USIDPRODt-1 + β6USFORRESt-1 + β7USGCINVt-1 + Ɛt..................Equation (12)

The Error Correction model is written as:

∆yt = β0 + β1∆ xt + (β3 – 1)( yt-1 - xt-1) + Ɛt................................................Equation (13)

The last term is the error correction term, reflecting disequilibrium responses. A major advantage of the VECM is that it is a restricted Vector Autoregressive System (VAR), with all the variables treated as endogenous. For the purpose of this research, the Error Correction Models are specified as:

## Country-Specific Variables- Nigeria indicators

∆NGNGDDt = β0 + β1∆NGNFDt + β2∆NGNCPIt-1 + β3∆NGNREAXt-1 + β4∆NGNIDPt-1 + β5∆NGNIMPt-1 + β6∆NGNIECt-1 + β7∆NGNCAPEXt-1 + β8∆NGNGOILRt-1 + (β8 – 1)( yt-1 - xt-1) + Ɛt ...........................................Equation (14)

∆NGNGEDt = β0 + β1∆NGNFDt + β2∆NGNCPIt-1 + β3∆NGNREAXt-1 + β4∆NGNIDPt-1 + β5∆NGNIMPt-1 + β6∆NGNIECt-1 + β7∆NGNCAPEXt-1 + β8∆NGNGOILRt-1 + (β8 – 1)( yt-1 - xt-1) + Ɛt .............................................Equation (15)

∆NGNGEDt = β0 + β1∆NGNFDt + β2∆NGNCPIt-1 + β3∆NGNREAXt-1 + β4∆NGNIDPt-1 + β5∆NGNIMPt-1 + β6∆NGNIECt-1 + β7∆NGNCAPEXt-1 + β8∆NGNGOILRt-1 + (β8 – 1)( yt-1 - xt-1) + Ɛt ...........................................Equation (16)

## External Factors Equations: US indicators

∆NGNGDDt = β0 + β1∆USCPIt + β2∆USGDPt-1 + β3∆USIMPORTSt-1 + β4∆USCAPEXt-1 + β5∆USIDPRODt-1 + β6∆USFORRESt-1 + β7∆USGCINVt-1 + (β8 – 1)( yt-1 - xt-1) + Ɛt....................Equation (17)

∆NGNGEDt = β0 + β1∆USCPIt + β2∆USGDPt-1 + β3∆USIMPORTSt-1 + β4∆USCAPEXt-1 + β5∆USIDPRODt-1 + β6∆USFORRESt-1 + β7∆USGCINVt-1 + (β8 – 1)( yt-1 - xt-1) + Ɛt ...................Equation (18)

∆NGNTBILLt = β0 + β1∆USCPIt + β2∆USGDPt-1 + β3∆USIMPORTSt-1 + β4∆USCAPEXt-1 + β5∆USIDPRODt-1 + β6∆USFORRESt-1 + β7∆USGCINVt-1 + (β8 – 1)( yt-1 - xt-1) + Ɛt .....................Equation (19)

Where;

## Country-specific financial soundness indicators;

NGNGDD= Nigerian Government Domestic Debt.

NGNGED= Nigerian Government External Debt.

NGNTBILL= Nigerian Interest Rate on Treasury Bills.

NGNFD= Nigerian Financial Deepening Index.

NGNCPI=Nigerian Consumer Price Index.

NGNREAX=Nigerian Real Exchange Rate (average of official exchange rate of the Naira vis a vis the United States Dollar).

NGNIDP=Nigerian Industrial Production index.

NGNIMP=Nigerian Mineral Production index.

NGNGOILR=Nigerian Government Oil Revenue.

NGNCAPEX=Nigerian Government Capital Expenditure.

NGNIEC=Nigerian Energy Consumption Index.

## External Factor financial soundness indicators;

USIDPROD=United States Industrial production Index.

USFORRES=United States Foreign Reserve Assets.

USGCINV=United States Government Consumption and Investment.

USCPI=United States Consumer Price Index

USGDP=United States Gross Domestic Product.

USIMPORTS=United States Imports of Goods and Services.

USCAPEX=United States Government Capital Expenditure.

USOUTHR=United States Output per Hour.

For the objectives of these research, following a conclusion that the times series data on the variables contains a unit root, the test for cointegration is employed using three main approaches.

## 3.8.1 Tests for Cointegration

## 3.8.1.1. i. Engle- Granger Approach

A simple method of testing cointegration, is the Engle and Granger test [58] , if the variables are integrated of order one [59] , we estimate with the OLS with the long run equilibrium equation:

Yt= β0 + β1X1t + β2X2t +.......+ βkXkt ....................Equation (20)

For the variables to be cointegrated the equilibrium errors must be stationary. The residuals of the regression are saved and tested for unit roots using the Augmented Dickey Fuller approach.

∆et= δet-1 + ∑ δ∆et-j+1 + vt ..................................................................Equation (21)

The Hypothesis is formed as follows:

H0: δ=0 (there is a unit root i.e. for no-cointegration).

H1: δ<1 (there is no unit root i.e. for cointegration).

If the residuals et are stationary, it means that the variables are cointegrated, or that there exists a long run equilibrium relationship among the variables.

## 3.8.1.2 ii. Durbin-Watson Approach

Another approach used for testing for cointegration, developed by Sargan and Bhargava (1983) is the Durbin-Watson approach. However it is similar in procedure to the Engle- Granger approach. We also estimate the equation using the OLS technique, save the residuals and compute the Durbin- Watson statistic on the residuals, which is now called the “cointegrating regression Durbin- Watson (CRDW) statistic.

CRDW= ∑ (et - et-1)2 where ē= arithmetic mean. ................................Equation (22)

∑ (et - ē)2

The hypothesis is formulated as follows:

H0 : Non-stationarity of et, (non cointegration, if CRDW < d

H1 : Stationarity of et, (cointegration, if CRDW > d.

If the CRDW is greater than the critical d values [60] , the null hypothesis is rejected, and we conclude that the variables are cointegrated.

## 3.8.1.3 iii. Johansen Cointegration Approach

In the situation where there are more than one cointegrating vectors, the Engle Granger methodology is no longer appropriate because it is assumes the existence of only one cointegrating vector. Therefore, cointegration among many variables the Johansen cointegration is more appropriate [61] . In line with the approach of Johansen Cointegration technique (1988, 1991), maximum eigenvalue and trace statistics will be employed to determine the number of cointegrating vectors when the variables are integrated of the same order. The major advantages of the Johansen approach is that it allows for testing of restrictions on the cointegrating relationship, and estimates more than one cointegrating vector in the equation.

The hypothesis is formulated as:

H0 :There is a no cointegrating vector/vectors among variables.

H1 : There is a cointegrating vector/vectors among variables.

Significant maximum eigenvalue and trace statistics implies that we reject the Null hypothesis of no cointegrating vectors. Johansen (1990), proposed two test statistics for the cointegrating ranks;

λ trace (r)= T ∑ In (1- λi ) ................................................................Equation (23)

λ max (r, r + 1)= T ∑ In (1- λi + 1 ) .....................................................Equation (24)

Where; λ trace is trace statistics, which is a likelihood ratio test statistic. λ max is the maximum eigenvalue statistic.

## 3.9 GRANGER CAUSALITY TESTS

An interesting phenomenon in time series analysis, is the Granger Causality, regression analysis focuses on the dependence of one variable on other variables, it does not represent causation or the direction of influence between the variables. In times series data analysis, it is very likely that one event causes another event. The Granger causality tests measures the direction of influence between the variables. There are three possible outcomes:

Unidirectional causality: is the case when Xt causes Yt, but Yt does not cause Xt. Bilateral or feedback causality: is the case when both variables Xt and Yt are jointly determined and Independence of the variables of each other.

A special definition of causality was developed by Granger (1969), as a variable x granger causes y if the prediction of the current value y is enhanced by employing past values of x. Empirical testing of causality involves regressing y on past, current and future values of x; if causality runs one way, from x to y, the set of coefficients of the future values of x should test insignificantly different from zero employing the F-test framework and the set of coefficients of the past values of x should test significantly different from zero. Prior to running this regression the data sets are tested for serial autocorrelation of the errors.

Yt =α10 + ∑ α1jXt-j + ∑ β1jYt-j + Ɛt.....................Equation (25)

Xt= α20 + ∑ α2jXt-j + ∑ β2jYt-j + Ɛ2t...................Equation (26)

Based on the model above, we can distinguish the following cases:

If (α11, α12,............, α1k)≠ 0 and (β21, β22,.............. β2k)=0, there exists a unidirectional causality from Xt to Yt , denoted as X→ Y.

If (α11, α12,............, α1k)=0 and (β21, β22,.............. β2k)≠0, there exists a unidirectional causality from Yt to Xt , denoted as Y→X.

If (α11, α12,............, α1k)≠0 and (β21, β22,.............. β2k)≠0, there exists a bilateral causality between Yt and Xt , denoted as X↔Y.

The Wald F-test framework is employed:

Fc = ....................................Equation (27)

Where:

SSRu = sum of squared residuals from the complete equation (unrestricted).

SSRr = sum of squared residuals from the complete equation, under the assumption that a set of variables is redundant. (restricted).

n= number of observations k=number of degrees of freedom.

The Hypothesis is formed as follows:

H0:X does not Granger cause Y, i.e. ((α11, α12,............, α1k)=0,if Fc< critical value of F

H1:X does Granger cause Y, i.e. ((α11, α12,............, α1k)≠0, if Fc > critical value of F

And

H0 :Y does not Granger cause X i.e. (β21, β22,.............. β2k)=0, if Fc< critical value of F

H1 :Y does Granger cause X, i.e. (β21, β22,.............. β2k)≠0, if Fc > critical value of F

## 3.10 VOLATILITY MODELLING IN TIME SERIES ANALYSIS

## 3.10.1 i. Tests for Autoregressive Conditional Heteroscedasticity (ARCH) Effects

To measure the volatility of financial and economic times series [62] ARCH models can be employed, because of the possibility that the variances are not constant over time. A very popular Arch model is the Generalized Autoregressive Conditional Heteroscedasticity [63] (GARCH). The simplest GARCH model is the GARCH (1,1) model, which is equivalent to the ARCH (2). Prior to estimating a GARCH type model, it is important to compute the tests for ARCH effects, for the purpose of ensuring that this class of models is appropriate for the data set . The process by which the variances are generated is assumed to be as follows; The pth-order ARCH process is;

σi2 = α0 + α1u 2t-1 + α2u 2t-2........+ αpu 2t-p.................Equation (28)

The hypothesis is formulated as:

H0 : α1 = 0 and α2 = 0 and .... αp =0.

H1 : α1 ≠ 0 or α2 ≠ 0 or........... αp ≠0.

If the value of the test statistic is greater than the critical value from the X2 (chi) distribution, we reject the null hypothesis of no arch effects.

## 3.10.2 ii. Estimating the Exponential Generalized Autoregressive Conditional Heteroscedasticity EGARCH

In estimating a GARCH type of model, the ordinary least squares is not appropriate since the model is non-linear, the maximum likelihood method is employed. The importance of ARCH models in finance cannot be overestimated because with financial time series it is very implausible that the variance of the error will be constant over time [64] . Thus ARCH models are very useful as they do not assume that the variance of the error is constant. For the purpose of testing the hypothesis of the volatility in BonnyLight high grade Nigerian oil prices, ARCH tests precede the GARCH/EGARCH tests, to ensure the approach is suitable for the data set.

GARCH(1,1), model is specified as;

…………..……………...Equation (31)

........................................Equation (30)

2

1

2

1

2

## '

##

##

##

##

##

##

##

##

##

t

t

t

t

t

t

r

x

y

##

##

##

##

##

##

##

σ 2t = conditional variance ω = mean

Ɛt-1 = lag of the squared residual from the mean equation (ARCH term).

σ 2t-1 = last period’s forecast variance (GARCH term).

The GARCH models are better and widely used than the ARCH model because it is more parsimonious and avoids over fitting [65] . However, the GARCH (1,1), imposes restrictions on α, β, and does not allow to test asymmetric effects of oil price shocks. The Exponential GARCH, (EGARCH) was proposed by Nelson (1991), following the methodology of Narayan and Narayan (2007), of modelling oil price volatility, The return on oil price is represented as:

...............................................................Equation (32)

Where; Pt= price of BonnyLight Nigerian Oil Pt-1 = price at the previous period.

On a continuous compounding basis, price is computed as the log of price at the end of the period less the log of price at the beginning of the period;

...................................................Equation (33)

Taking the Logarithm of price as a random walk process;

.............................................................Equation (34)

This is the same as;

.................................Equation (35)

where Ɛt = σωt and ωt ~ IID N(0,1).

IID= Independently and Identically Distributed.

For the Exponential GARCH- EGARCH (1,1) in the mean model, the mean and variance structures are as follows;

## Mean Equation:

.....................Equation (36)

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

##

## p

## j

## q

## i

## i

## t

## i

## t

## i

## t

## i

## t

## i

## j

## t

## j

## t

## y

## 1

## 1

## 2

## 2

## log

## log

##

##

##

##

##

##

##

##

## Variance Equation:

.....................Equation (37)

The EGARCH is employed because it does not impose any restrictions on α, y and β, also asymmetries are allowed under the formulation, which is captured by the parameters y, if y>0, the implication is that positive shocks gives rise to higher volatility than negative shocks. The estimate of β also allows for the evaluation of whether shocks to the variance are persistent or not. Representing the magnitude of the conditional shock on the conditional variance is α. To obtain robust inference about the estimated models, the robust standard errors are computed by Bollerslev-Wooldridge and as suggested by Narayan and Narayan (2007). The EGARCH (1,1), is estimated using the maximum likelihood estimation technique assuming normally distributed errors and the optimal lag length selected using the Schwartz Bayesian criterion.

## 3.10.3 STABILITY AND RESIDUAL TESTS;

A common problem with time series data is serial correlation, which simply means correlation between members of series of observations ordered in time. For this research, the Breusch Godfrey test will be employed to test for serial correlation. Another important test is specification error, however when testing for cointegration if a relevant variable has been omitted from the equation, the test should fail to find cointegration among the variables [66] . In addition, an important assumption with the vector autoregressive system, is that all the variables in the system are endogenous.

Therefore, there is no need to test for Multicolinearity, as it is an assumption based on exogenous variables. All estimations are carried out using Bollerslev-Wooldridge Standard errors and Newey West Standard Errors, is also employed to correct for serial correlation and Heteroscedasticity, it’s an extension of White’s Heteroscedasticity Consistent Standard Errors.

## 3.11.1 DATA EVALUATION TECHNIQUES

## 3.11.1.1. P-values

This is commonly known as the exact significance level or the probability of being wrong when the null hypothesis is rejected. [67] The p-value can also be termed the plausibility of the null hypothesis, so the smaller the p-value the less plausible is the null hypothesis. It is also known as the exact probability of committing type 1 error [68] .

## 3.11.1.2 Test of Significance

This is used to estimate the parameters at a certain level of significance. The [69] test of significance procedure involves verifying the truth or falsity of a null hypothesis [70] . It determines the region where the null hypothesis under the test will be rejected or not rejected. If the t-computed is greater than that obtained from the table, we conclude that our estimate is significant and hence reject the null hypothesis.

T- Ratio is given by: t* =

Se () ............................................................Equation (38)

where Se= standard error β=coefficient estimate.

## 3.11.1.3 F-test Statistics

This framework is employed to evaluate the validity of the hypothesis on the overall significance of the model to be postulated in the study. The [71] F-test involves two regressions, the restricted regression in which coefficients are restricted and the unrestricted regressions in which coefficients are freely determined by the data. F-distribution has only positive values. The f – ratio is given by:

F* = R²/ (k-1)

(1-R²)/(n-k) ....................................................................................Equation (39)

Where: k = number of estimated parameters n= number of observations R²= co-efficient of determination

## 3.11.1.4 Co-efficient of Determination (R2)

This is commonly referred to as the measure of “goodness of fit”. This is a summary measure that tells how well the sample regression line fits the data. Precisely, it measures the percentage of the total variation in the dependent variable explained by the regression model. There are two major properties of R2 it’s a non-negative quantity and its limit are between 0 and 1. An R2 of 1 implies a perfect fit and a 0 value implies there is no relationship between the dependent and independent variables [72] .

R²= ESS or 1 - RSS

TSS TSS...............................................................Equation (40)

Where; ESS- Explained Sum of Squares- ∑ (Ŷi - Ȳ)2

RSS- Residual Sum of Squares- ∑ ûi2

TSS- Total Sum of Squares- ∑ (Yi - Ȳ)2

## 3.12 SAMPLE RATIONALE

For the purpose of this research, the sample represents quarterly observations obtained from Central bank of Nigeria [73] , and National Bureau of Statistics [74] for the country specific variables and from DataStream for the external factors variables [75] . The quarterly observations is from the period 1970 to 2009. In year 2007, the financial crisis started in the United States, with the sub-prime mortgage crisis, however this crisis has been termed as a global crisis affecting major economies of the world. In 1970, the Nigerian Civil War [76] which started in 1967, came to an end. It caused instability in the financial and economic sectors, resulting in banking crisis, shutdown of various businesses, and loss of jobs, this lead to various channels of financial leverage obtained by the Government.

At the end of the Civil War, and at the beginning of 1980s, many emerging economies including Nigeria embraced financial sector reforms to boost the development of their economies. Financial liberalization is generally defined as the abolition of credit control, deregulation of interest rates, ease of entry into the financial services industry, development of capital markets, and liberalization of international capital flows. The Debt Management Office of the Nigerian Government, reported that the extent of leverage has increased over the years from the 1980s to the 2000s due to financial liberalization. Therefore, the period of study is essential to determining the effects of certain financial and economic factors on macroeconomic financial leverage.

For the purpose of modelling oil price volatility in Nigeria, the BonnyLight High Grade Nigerian Oil price is employed, and 2504 daily observations is obtained from DataStream from the period 30/11/2000 to 07/07/2010. The period of analysis is split into 2 sub periods, on the basis of examining the pre-financial crisis, global oil increase, the global warming increase and the financial crisis period. The Global financial crisis [77] has been argued to impact the Nigerian economy, through the volatility of oil price because it’s an oil dependent economy. According to the International Monetary Fund (IMF), in November 2000 the world oil price averaged over three times higher than its February 1999 price and as such a global oil increase was declared.

The Global warming, a major environmental threat caused by excessive build-up of heat trapping “greenhouse” gases in the Earth’s atmosphere was argued to push oil prices causing volatility in oil price. In 2006, Global warming reports indicated that the effects of the environmental threat increased drastically. Hence, the pre- financial crisis and global oil increase period in this study is from 30th of November 2000 to 30th of January 2006, the global warming increase and financial crisis period is from 1st of February 2006 to 7th of July, 2010.

## CHAPTER FOUR

## DATA PRESENTATION, INTERPRETATION AND ANALYSIS

## 4.1 Introduction

In the previous Chapter, the research methodology of this study was discussed, in terms of the various tests to be carried out inorder to achieve its research objectives and tests it research hypotheses. Therefore, this chapter focuses on presenting, analyzing the results and interpretation of the data employed in this study.

The results of the Descriptive Statistics are first presented, followed by the Serial Correlation Tests and Stability tests. The results of the Stationarity tests by Augmented Dickey Fuller, Philips Perron, and Kwiatkowski-Phillips-Schmidt test are also presented for each Country Specific Variable and External Factor Variable, followed by the results of the Cointegration by Engle Granger, Durbin Watson and Johansen techniques for each Country Specific and External Factor Equation. The Error Correction Estimates are presented for each Country Specific and External Factor Equation also followed by the Granger Causality results. The results of the Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH) tests are also presented with the tests for Autoregressive Conditional Heteroscedasticity (ARCH) effects.

## 4.2 DESCRIPTIVE STATISTICS

The descriptive statistics is presented below for each Country Specific and External Factor indicators.

## 4.2.1 Country Specific Financial Soundness Indicators

## Table 1- Descriptive Statistics

## Mean

## Median

## Max

## Min

## Std. Dev.

## NGNTBILL

## 0.10

## 0.085

## 0.296

## 0.025

## 0.060802

## NGNFD

## 23.48

## 20.25

## 40.2

## 9.2

## 9.324773

## NGNGDD

## 53.64

## 62.1

## 299786

## 1000.7

## 8584.99

## NGNGOILR

## 12.52

## 29.6

## 590287

## 166.6

## 1655901

## NGNREAX

## 38.39

## 7.714

## 133.50

## 0.5464

## 53.03354

## NGNIDP

## 58.4

## 126.9

## 163.4

## 41.3

## 31.82742

## NGNIMP

## 78.8

## 77.95

## 189

## 72.2

## 28.02745

## NGNCPI

## 77.5

## 82.95

## 9789.1

## 10.8

## 3119.077

## NGNCAPEX

## 6.682

## 4.85

## 20

## 2.4

## 4.561314

## NGNIEC

## 92.7

## 96.3

## 177.1

## 21.5

## 49.82768

## NGNGED

## 75.1

## 64.1

## 489027

## 175

## 131.2932

Relatively High Leverage: An average leverage ratio of approximately 75.1% suggests that Nigerian Government level of macroeconomic financial leverage is high as reported by the Debt Management Office of the Nigerian Government.

Relatively High Industrial and Mineral Production: An average ratio of approximately 58.4% suggests a high level of industrial production and 78.8% suggests high level of mineral production in Nigeria.

Relatively High Consumer Price Index: An average ratio of approximately 77.5% suggests a high level of consumer price index.

Relatively High Energy Consumption: An average ratio of approximately 92.7% suggests a high level of energy consumption. This is consistent with previous studies on oil exporting and emerging markets economy.

## 4.2.2 External Factors Financial Soundness Indicators

## Table 2- Descriptive Statistics

## Mean

## Median

## Max

## Min

## Std. Dev.

## USCPI

## 27.01

## 30.25

## 214.774

## 39.8

## 54.44

## USIMPORTS

## 94.05

## 58.35

## 2188

## 240.2

## 652.24

## USCAPEX

## 51.76

## 68.77

## 1178.3

## 114.0

## 316.68

## USIDPROD

## 65.32

## 61.61

## 100.47

## 37.03

## 19.722

## USGDP

## 86.5

## 76.8

## 13391.2

## 4256.6

## 28.9997

## USFORRES

## 59.2

## 65.1

## 86824

## 12315

## 26.683.22

## USGCINV

## 84.8

## 79.2

## 2955.4

## 238.8

## 790.86

## USOUTHR

## 50.03

## 94.05

## 150.9

## 67.06

## 24.045

Relatively High Level of Imports : An average leverage ratio of approximately 94.05% suggests that United States Government has a high level of imports.

Relatively High Industrial Production: An average ratio of approximately 65.3% suggests a high level of industrial production in the United States.

Relatively High Consumer Price Index: An average ratio of approximately 27.01% suggests a high level of consumer price index in United States.

Relatively High Foreign Reserve Assets: An average ratio of approximately 59.2% suggests a high level of Foreign Reserve Assets in the United States.

Relatively High Government Consumption and Investment: An average ratio of approximately 84.8% suggests a high level of Government Consumption and Investment in the United States.

## 4.3 SERIAL CORRELATION TESTS

The Breusch-Godfrey Serial Correlation LM test is employed to test higher orders of serial correlation, the AR(10) structure in this study.

## 4.3.1 Country Specific Financial Soundness Indicators

The results of the BG test are presented below:

## Table 3- Serial Correlation tests for Country Specific equations

## Country Specific Equations

## F-Statistics

## Probability

## Equation 7

## 4.612329*

## 0.0182

## Equation 8

## 4.993406*

## 0.0017

## Equation 9

## 0.466139

## 0.6320

## *, * *, *** represents significance level at 1,5 and 10% respectively. Equation 1,2,3 are in page

The Null and Alternative hypothesis are;

HA: ρ1 = ρ2 = ρ3 = ρ4 = ρ5 = ρ6 = ρ7 = ρ8 = ρ9 = ρ10 = 0........................AR (10)

HB : ρ1 ≠ ρ2 ≠ ρ3 ≠ ρ4 ≠ ρ5 ≠ ρ6 ≠ ρ7 ≠ ρ8 ≠ ρ9 ≠ ρ10 ≠ 0..................AR (10)

The Null hypothesis of no serial correlation is rejected at 1% level of significance, for Country Specific Equations 1 and 2, as presented above.

## 4.3.2 External Factors Financial Soundness Indicators

The results of the BG test are presented below:

## Table 4- Serial Correlation tests for External Factors equations

## External Factors Equations

## F-Statistics

## Probability

## Equation 10

## 1.809490

## 0.1817

## Equation 11

## 1.426323

## 0.2565

## Equation 12

## 2.259449

## 0.1225

## *, * *, *** represents significance level at 1,5 and 10% respectively. Equation 4,5,6 are in page

The Null and Alternative hypothesis are;

HA: ρ1 = ρ2 = ρ3 = ρ4 = ρ5 = ρ6 = ρ7 = ρ8 = ρ9 = ρ10 = 0........................AR (10)

HB : ρ1 ≠ ρ2 ≠ ρ3 ≠ ρ4 ≠ ρ5 ≠ ρ6 ≠ ρ7 ≠ ρ8 ≠ ρ9 ≠ ρ10 ≠ 0..................AR (10)

The Null hypothesis of no serial correlation is not rejected at 1,5 or 10% level of significance for External Factor Equations 4,5 and 6 as presented above.

## 4.2.3 Newey- West Standard Errors

The Newey- West method used to obtain standard errors of Ordinary least estimators that are corrected for serial correlation. This method is an extension of the White’s Heteroscedasticity- consistent standard errors methodology. To correct for serial correlation the Newey- West standard errors are employed in this study.

## 4.3 TESTS FOR STATIONARITY

## 4.3.1 Country Specific Financial Soundness Indicators

The Augmented Dickey Fuller, Philips Perron, and Kwiatkowski-Phillips-Schmidt tests are employed to test for stationarity following the Box Jenkins methodology to modelling time series. The use of non-stationarity data can lead to spurious regressions. Inorder to avoid this the tests are employed. The results are below;

## 4.3.1.1. Augmented Dickey Fuller and Philips Perron

## Table 5- ADF and PP tests

## Country Specific

## ADF

## PP

## Without Trend

## With Trend

## Without Trend

## With Trend

## Variables

## Level

## FD

## Level

## FD

## Level

## FD

## Level

## FD

## NGNTBILL

## -2.17

## -7.21*

## -2.09

## -6.04*

## -2.09

## -7.59*

## -1.91

## -8.78*

## NGNGDD

## -1.45

## -2.74**

## -3.15

## -4.62*

## -2.02

## -4.86*

## -3.12

## -6.24*

## NGNGED

## 2.08

## -3.99**

## 0.53

## -12.1*

## -1.98

## -5.96*

## -2.22

## -5.92*

## NGNFD

## -1.34

## -5.58*

## -1.21

## -5.51*

## -1.52

## -5.59*

## -1.39

## -5.54*

## NGNGOILR

## -2.23

## -3.66**

## -3.12

## -4.57*

## -2.09

## -14.9*

## -3.6

## -16.5*

## NGNREAX

## 0.36

## -5.28*

## 5.28

## -5.47*

## 0.22

## -5.29*

## -1.49

## -5.47*

## NGNIDP

## -2.18

## -6.85*

## -3.39

## -6.88*

## -2.22

## -6.85*

## -3.38

## -6.88*

## NGNIMP

## -2.23

## -7.54*

## -3.12

## -7.42*

## -2.91

## -8.15*

## -3.26

## -8.17*

## NGNCPI

## 1.45

## -2.71**

## 0.84

## -3.91*

## 4.32

## -2.27**

## 0.57

## -3.87**

## NGNCAPEX

## -1.71

## -9.44*

## -2.66

## -9.46*

## -2.57

## -9.29*

## -3.22

## -9.48*

## NGNIEC

## -1.39

## -5.45*

## -2.17

## -5.37*

## -1.38

## -5.44*

## -2.17

## -5.34*

## Notes: ADF- Augmented Dickey Fuller, PP- Philips Perron, FD- First Difference *, * *, *** represents significance level at 1,5 and 10% respectively.

The Null and Alternative hypothesis:

HA: δ=0 (there is a unit root or the time series is nonstationary tδ > T).

HB: δ<1 (there is no unit root or the time series is stationary tδ < T).

The computed ADF/PP statistics do not exceed the ADF/PP critical tau values, so we do not reject the null hypothesis of unit root, however as reported in Table 1 presenting the ADF and PP tests, the time series are stationary after taking first difference for all the country specific variables at significance level of 1, 5 and 10%, therefore we reject the Null hypothesis of a Unit root after taking the first differences of each country specific indicator. This is consistent with the Box Jenkins methodology and previous empirical research by Ayadi et. al., (2002, 2008), Olomola and Adejumo (2006) and Omisakin et. al., (2009).

## 4.3.1.2 Kwiatkowski-Phillips-Schmidt test

The Null and Alternative hypothesis for the KPSS test are a reverse of the ADF and PP tests for stationarity. The hypothesis are below;

HA: δ<1 (there is no unit root or the time series is stationary tδ < T).

HB: δ=0 (there is a unit root or the time series is nonstationary tδ > T).

The computed KPSS statistics exceed the KPSS critical tau values, so we reject the null hypothesis of no unit root, however as reported in Table 6 presenting the KPSS tests, the time series are stationary after taking first difference for all the country specific variables at significance level of 1, 5 and 10%, therefore we do not reject the Null hypothesis of a no Unit root after taking the first differences. The KPSS results also further confirmed that all the series are stationary after first differencing, consistent with the ADF, PP, and Box Jenkins methodology and consistent with previous empirical research by Ayadi et. al., (2002, 2008), Olomola and Adejumo (2006) and Omisakin et. al., (2009).

## Table 6- KPSS tests

## Country Specific Variables

## KPSS: Kwiatkowski-Phillips-Schmidt-Shin test statistic

## Without Trend

## With Trend

## Variables

## Level

## FD

## Level

## FD

## NGNTBILL

## 0.3664***

## 0.2087

## 0.1786**

## 0.1457

## NGNGDD

## 0.6099**

## 0.5923

## 0.2182*

## 0.1462

## NGNGED

## 0.4578***

## 0.0782

## 0.3161***

## 0.0741

## NGNFD

## 0.4731**

## 0.1659

## 0.1565**

## 0.1192

## NGNGOILR

## 0.7201**

## 0.5001

## 0.1102***

## 0.0082

## NGNREAX

## 0.6118**

## 0.2632

## 0.1772**

## 0.0759

## NGNIDP

## 0.7487*

## 0.1566

## 0.1465**

## 0.0595

## NGNIMP

## 0.4731**

## 0.0064

## 0.1319***

## 0.0667

## NGNCPI

## 0.6384**

## 0.2475

## 0.2022**

## 0.1147

## NGNCAPEX

## 0.3701***

## 0.1095

## 0.1461**

## 0.0441

## NGNIEC

## 0.5675**

## 0.0704

## 0.7211**

## 0.0699

## Notes: KPSS: Kwiatkowski-Phillips-Schmidt-Shin test statistic , FD- First Difference *, * *, *** represents significance level at 1,5 and 10% respectively.

## 4.3.2 External Factors Financial Soundness Indicators

The Augmented Dickey Fuller, Philips Perron, and Kwiatkowski-Phillips-Schmidt tests are employed to test for stationarity following the Box Jenkins methodology to modelling time series. The results are below;

## 4.3.2.1. Augmented Dickey Fuller and Philips Perron

The Null and Alternative hypothesis:

HA: δ=0 (there is a unit root or the time series is nonstationary tδ > T).

HB: δ<1 (there is no unit root or the time series is stationary tδ < T).

The computed ADF/PP statistics do not exceed the ADF/PP critical tau values, so we do not reject the null hypothesis of unit root, however as reported in Table 7 presenting the ADF and PP tests, the time series are stationary after taking first difference for all the external factors variables at significance level of 1, 5 and 10%, therefore we reject the Null hypothesis of a Unit root after taking the first differences. This is consistent with the Box Jenkins methodology and previous empirical research by Ayadi et. al., (2002, 2008), Olomola and Adejumo (2006) and Omisakin et. al., (2009).

## Table 7-ADF and PP tests

## External Factors Variables

## ADF

## PP

## Without Trend

## With Trend

## Without Trend

## With Trend

## Variables

## Level

## FD

## Level

## FD

## Level

## FD

## Level

## FD

## USCPI

## -0.06

## -4.96*

## -2.62

## -4.88*

## -0.06

## -4.92*

## -2.42

## -4.84*

## USIMPORTS

## -1.43

## -3.05**

## -2.14

## -3.19**

## -0.04

## -3.05**

## -1.82

## -4.29*

## USCAPEX

## -0.14

## -3.99*

## -2.34

## -3.93**

## -0.78

## -2.97**

## -2.02

## -4.32*

## USGDP

## -0.21

## -3.74**

## -2.28

## -3.35**

## 0.33

## -3.32**

## -1.91

## -3.53**

## USFORRES

## -1.18

## -5.79*

## -2.55

## 5.78*

## -0.86

## -7.87*

## -2.03

## -9.17*

## USGCINV

## 2.12

## -2.91**

## -0.42

## -3.28**

## 4.92

## -3.61**

## 0.55

## -3.32**

## USOUTHR

## -2.17

## -4.55*

## 0.22

## -5.48*

## 3.13

## -4.74*

## 0.18

## -5.48*

## USIDPROD

## -1.16

## -4.07*

## -2.22

## -4.07*

## -1.09

## -4.04*

## -1.66

## -4.05**

## Notes: ADF- Augmented Dickey Fuller, PP- Philips Perron, FD- First Difference *, * *, *** represents significance level at 1,5 and 10% respectively.

## 4.3.2.2 Kwiatkowski-Phillips-Schmidt test

The Null and Alternative hypothesis for the KPSS test are a reverse of the ADF and PP tests for stationarity. The hypothesis are below;

HA: δ<1 (there is no unit root or the time series is stationary tδ < T).

HB: δ=0 (there is a unit root or the time series is nonstationary tδ > T).

The computed KPSS statistics exceed the KPSS critical tau values, so we reject the null hypothesis of no unit root, however as reported in Table 8 presenting the KPSS tests, the time series are stationary after taking first difference for all the external factors variables at significance level of 1, 5 and 10%, therefore we do not reject the Null hypothesis of a no Unit root after taking the first differences. The KPSS results also further confirmed that all the series are stationary after first differencing, consistent with the ADF, PP, and Box Jenkins methodology.

## External Factors Variables

## KPSS: Kwiatkowski-Phillips-Schmidt-Shin test statistic

## Without Trend

## With Trend

## Variables

## Level

## FD

## Level

## FD

## USCPI

## 0.7754*

## 0.0835

## 0.1257***

## 0.0854

## USIMPORTS

## 0.7072**

## 0.1913

## 0.1845**

## 0.1248

## USCAPEX

## 0.7155**

## 0.0864

## 0.1235***

## 0.0898

## USGDP

## 0.7618***

## 0.2482

## 0.1834**

## 0.1039

## USFORRES

## 0.6862**

## 0.0401

## 0.1695**

## 0.5001

## USGCINV

## 0.7591*

## 0.5412

## 0.1768**

## 0.1194

## USOUTHR

## 0.7538*

## 0.5585

## 0.1913**

## 0.0097

## USIDPROD

## 0.7397*

## 0.1301

## 0.1334***

## 0.1175Table 8- KPSS tests

## Notes: KPSS: Kwiatkowski-Phillips-Schmidt-Shin test statistic , FD- First Difference *, * *, *** represents significance level at 1,5 and 10% respectively.

## 4.4 MODELLING LONG RUN RELATIONSHIPS IN FINANCE

## 4.4.1 Macroeconomic Financial Leverage and Country Specific Financial Soundness Indicators.

To test the present of a long run equilibrium relationship, The Engle Granger, Durbin Watson and Johansen Cointegration Technique is employed in this study.

## 4.4.1.1. Engle Granger and Durbin Watson Cointegration

The results in Table 5 presents the Engle Granger and Dubin Watson test for cointegration for the Country Specific equations. The residuals of the equations are stationary for the Engle Granger test, and the CRDW is greater than the d critical value indicating that the null hypothesis of no cointegration is rejected for both tests at 1% level of significance for all cases. Therefore a long run equilibrium relationship exists among macoreconomic financial leverage and the country specific financial soundness indicators which are Industrial production, Mineral production, Energy Consumption, Consumer Price, and Financial Deepening Indexes, Government Oil revenue, Real exchange rate, and Government Capital Expenditure. The Figure 1.3 [78] also presents the residuals of the equations stationarity.

## Table 9- EG and DW tests

## Country Specific Equations

## ENGLE GRANGER COINTEGRATION

## DURBIN WATSON COINTEGRATION

## ADF t-stat

## p-value

## CRDW

## d-value

## Equation 7

## -3.984*

## 0.0037

## Equation7

## 2.168*

## 0.511

## Equation 8

## -5.277*

## 0.0001

## Equation 8

## 2.434*

## 0.511

## Equation 9

## -5.729*

## 0.0000

## Equation 9

## 2.031*

## 0.511

## Notes: ADF- Augmented Dickey Fuller, CRDW- Cointegrating Regression Durbin Watson,

## d-value- critical values computed by Sargan and Bhargava (1983), Engle and Granger (1987) [79]

## *, * *, *** represents significance level at 1,5 and 10% respectively. Equation 1,2,3 are in page

## 4.4.1.2 Johansen Cointegration

The Johansen Cointegration technique is also employed, the results of each country specific equation is presented in Tables 10, 11, and 12. The results of the Trace and Maximum Eigen value statistic indicates at most 9 cointegrating equations at 5% level of significance for all cases. Therefore the null hypothesis of no cointegration is rejected for all the equations. Therefore a long run equilibrium relationship exists among macroeconomic financial leverage and the country specific financial soundness indicators, confirming the results of the Engle Granger and Durbin Watson Cointegration.

## Table 10- Johansen Cointegration- Equation 7

## JOHANSEN COINTEGRATION TECHNIQUE

## Country Specific Variables- Equation 7

## Hypothesized No. of CE (s)

## Max-Eigen Statistic

## Critical Value (Eigen) at 5%

## Trace Statistic

## Critical Value (Trace) at 5%

## None*

## 239.68

## 58.433

## 697.77

## 197.37

## At most 1*

## 130.69

## 52.362

## 458.08

## 159.52

## At most 2*

## 99.935

## 46.231

## 327.39

## 125.61

## At most 3*

## 74.379

## 40.077

## 227.46

## 95.75

## At most 4*

## 60.404

## 33.876

## 153.08

## 69.81

## At most 5*

## 38.438

## 27.584

## 92.67

## 47.85

## At most 6*

## 22.321

## 21.131

## 54.237

## 29.99

## At most 7*

## 18.022

## 14.264

## 31.917

## 15.49

## At most 8*

## 13.895

## 3.814

## 13.895

## 3.841

## Trace and Max-eigenvalue test indicates 9 cointegrating eqn(s) at the 0.05 level

## * denotes rejection of the hypothesis at the 0.05 level.

## Table 11- Johansen Cointegration- Equation 8

## JOHANSEN COINTEGRATION TECHNIQUE

## Country Specific Variables- Equation 8

## Hypothesized No. of CE (s)

## Max-Eigen Statistic

## Critical Value (Eigen) at 5%

## Trace Statistic

## Critical Value (Trace) at 5%

## None*

## 297.62

## 58.433

## 804.542

## 197.37

## At most 1*

## 140.73

## 52.362

## 506.917

## 159.52

## At most 2*

## 125.26

## 46.231

## 366.185

## 125.61

## At most 3*

## 97.914

## 40.077

## 240.923

## 95.75

## At most 4*

## 52.393

## 33.876

## 143.009

## 69.81

## At most 5*

## 42.145

## 27.584

## 90.615

## 47.85

## At most 6*

## 22.933

## 21.131

## 48.669

## 29.79

## At most 7*

## 16.312

## 14.264

## 25.536

## 15.49

## At most 8*

## 9.223

## 3.841

## 9.2234

## 3.841

## Trace and Max-eigenvalue test indicates 9 cointegrating eqn(s) at the 0.05 level

## * denotes rejection of the hypothesis at the 0.05 level.

## -

## JOHANSEN COINTEGRATION TECHNIQUE

## Country Specific Variables- Equation 9

## Hypothesized No. of CE (s)

## Max-Eigen Statistic

## Critical Value (Eigen) at 5%

## Trace Statistic

## Critical Value (Trace) at 5%

## None*

## 141.45

## 58.43

## 500.57

## 197.37

## At most 1*

## 103.94

## 52.36

## 359.11

## 159.52

## At most 2*

## 73.257

## 46.23

## 255.17

## 125.61

## At most 3*

## 61.972

## 40.07

## 181.91

## 95.75

## At most 4*

## 48.591

## 33.87

## 119.93

## 69.81

## At most 5*

## 29.542

## 27.58

## 71.34

## 47.85

## At most 6*

## 23.471

## 21.13

## 41.79

## 29.79

## At most 7*

## 17.089

## 14.26

## 18.32

## 15.49

## At most 8

## 1.2365

## 3.841

## 1.236

## 3.841

## Trace and Max-eigenvalue test indicates 8 cointegrating eqn(s) at the 0.05 level

## * denotes rejection of the hypothesis at the 0.05 levelTable 12- Johansen Cointegration- Equation 9

Therefore the following research hypothesises are not rejected at either 1,5 or 10% level of significance;

H1: There is a long run equilibrium relationship between Nigerian Government Oil revenue and macroeconomic financial leverage in Nigeria.

H2: There is a long run equilibrium relationship between Nigerian Financial Deepening Index and macroeconomic financial leverage in Nigeria.

H3: There is a long run equilibrium relationship between Nigerian Capital Expenditure and macroeconomic financial leverage in Nigeria.

H4: There is a long run equilibrium relationship between Real exchange rate (NGN/US) [80] and macroeconomic financial leverage in Nigeria.

H5: There is a long run equilibrium relationship between Energy Consumption Index and macroeconomic financial leverage in Nigeria.

H6: There is a long run equilibrium relationship between Mineral Production Index and macroeconomic financial leverage in Nigeria.

H7: There is a long run equilibrium relationship between Consumer Price Index and macroeconomic financial leverage in Nigeria.

H8: There is a long run equilibrium relationship between Nigerian Industrial Production and macroeconomic financial leverage in Nigeria.

H9: There is a long run equilibrium relationship between the Nigerian Financial Soundness Indicators and macroeconomic financial leverage in Nigeria.

## 4.4.2 Macroeconomic Financial Leverage and External Factors Financial Soundness Indicators.

To test the present of a long run equilibrium relationship, The Engle Granger, Durbin Watson and Johansen Cointegration Technique is employed in this study.

## 4.4.2.1 Engle Granger and Durbin Watson Cointegration

The results in Table 9 presents the Engle Granger and Dubin Watson test for cointegration for the external factors equations. The residuals of the equations are stationary for the Engle Granger test, and the CRDW is greater than the d critical value indicating that the null hypothesis of no cointegration is rejected for both tests at 1% level of significance for all cases. Therefore a long run equilibrium relationship exists among macoreconomic financial leverage and the external factors financial soundness indicators which are Industrial production and Consumer Price Indexes, Foreign Reserve Assets, Government Consumption and Investment, Gross Domestic Product, Imports of Goods and Services, Government Capital Expenditure and Output per Hour.

## Table 13 – EG and DW tests

## External Factors Equations

## ENGLE GRANGER COINTEGRATION

## DURBIN WATSON COINTEGRATION

## ADF t-stat

## p-value

## CRDW

## d-value

## Equation 10

## -4.724*

## 0.0005

## Equation 10

## 1.961*

## 0.511

## Equation 11

## -5.892*

## 0.0000

## Equation 11

## 2.012*

## 0.511

## Equation 12

## -5.397*

## 0.0000

## Equation 12

## 2.125*

## 0.511

## Notes: ADF- Augmented Dickey Fuller, CRDW- Cointegrating Regression Durbin Watson, d-value- critical values computed by Sargan and Bhargava (1983), Engle and Granger (1987) [81] *, * *, *** represents significance level at 1,5 and 10% respectively. Equation 4,5,6 are in page

## 4.4.2.2 Johansen Cointegration

The Johansen Cointegration technique is also employed, the results of each external factor equation is presented in Tables 14,15, and 16. The results of the Trace and Maximum Eigen value statistic indicates at most cointegrating equations at 5% level of significance for all cases. Therefore the null hypothesis of no cointegration is rejected for all the equations. Therefore a long run equilibrium relationship exists among macroeconomic financial leverage and the external factors financial soundness indicators, confirming the results of the Engle Granger and Durbin Watson Cointegration.

## Table 14- Johansen Cointegration- Equation 10

## JOHANSEN COINTEGRATION TECHNIQUE

## External Factor Variables Equation 10

## Hypothesized No. of CE (s)

## Max-Eigen Statistic

## Critical Value (Eigen) at 5%

## Trace Statistic

## Critical Value (Trace) at 5%

## None*

## 144.24

## 52.362

## 387.15

## 159.52

## At most 1*

## 81.162

## 46.231

## 242.91

## 125.61

## At most 2*

## 51.778

## 40.077

## 161.74

## 95.758

## At most 3*

## 45.123

## 33.876

## 109.97

## 69.818

## At most 4*

## 27.764

## 27.584

## 64.846

## 47.856

## At most 5*

## 19.021

## 21.131

## 37.082

## 29.747

## At most 6*

## 16.874

## 14.264

## 18.062

## 13.494

## At most 7

## 1.1879

## 3.814

## 1.1879

## 3.8414

## Trace test and Max-eigenvalue test indicates 7 and 5 cointegrating eqn(s) respectively at the 0.05 level

## * denotes rejection of the hypothesis at the 0.05 level.

## Table 15- Johansen Cointegration- Equation 11

## JOHANSEN COINTEGRATION TECHNIQUE

## External Factor Variables Equation 11

## Hypothesized No. of CE (s)

## Max-Eigen Statistic

## Critical Value (Eigen) at 5%

## Trace Statistic

## Critical Value (Trace) at 5%

## None*

## 164.26

## 58.433

## 605.52

## 197.37

## At most 1*

## 122.53

## 52.362

## 441.29

## 159.52

## At most 2*

## 97.546

## 46.231

## 318.76

## 125.61

## At most 3*

## 82.703

## 40.077

## 221.21

## 95.753

## At most 4*

## 59.162

## 33.871

## 138.51

## 69.818

## At most 5*

## 35.741

## 27.131

## 79.347

## 47.856

## At most 6*

## 25.142

## 21.131

## 43.607

## 29.797

## At most 7*

## 13.366

## 14.264

## 18.464

## 15.494

## At most 8*

## 5.0984

## 3.841

## 5.0984

## 3.841

## Trace test and Max-eigenvalue test indicates 9 and 7 cointegrating eqn(s) respectively at the 0.05 level

## * denotes rejection of the hypothesis at the 0.05 level.

## Table 16 Johansen Cointegration- Equation 12

## JOHANSEN COINTEGRATION TECHNIQUE

## External Factor Variables Equation 12

## Hypothesized No. of CE (s)

## Max-Eigen Statistic

## Critical Value (Eigen) at 5%

## Trace Statistic

## Critical Value (Trace) at 5%

## None*

## 371.173

## 58.433

## 822.749

## 197.37

## At most 1*

## 138.605

## 52.362

## 451.596

## 159.509

## At most 2*

## 94.784

## 46.231

## 312.919

## 125.615

## At most 3*

## 70.826

## 40.077

## 218.186

## 95.753

## At most 4*

## 64.055

## 38.876

## 149.36

## 69.818

## At most 5*

## 35.371

## 27.584

## 83.304

## 47.856

## At most 6*

## 28.774

## 21.131

## 47.933

## 29.747

## At most 7*

## 15.546

## 14.264

## 19.154

## 15.494

## At most 8*

## 3.608

## 3.8414

## 3.608

## 3.8417

## Trace and Max-eigenvalue test indicates 8 cointegrating eqn(s) at the 0.05 level

## * denotes rejection of the hypothesis at the 0.05 level.

Therefore the following research hypothesises are not rejected at either 1,5 or 10% level of significance;

H10: There is a long run equilibrium relationship between the United States Gross Domestic Product and macroeconomic financial leverage in Nigeria.

H11: There is a long run equilibrium relationship between the United States Foreign reserve assets and macroeconomic financial leverage in Nigeria.

H12: There is a long run equilibrium relationship between United States Consumer Price Index and macroeconomic financial leverage in Nigeria.

H13: There is a long run equilibrium relationship between United States Imports of goods and services and macroeconomic financial leverage in Nigeria.

H14: There is a long run equilibrium relationship between United States Industrial production and macroeconomic financial leverage in Nigeria.

H15: There is a long run equilibrium relationship between United States Government Consumption and Investment and macroeconomic financial leverage in Nigeria.

H16: There is a long run equilibrium relationship between United States Government Capital Expenditure and macroeconomic financial leverage in Nigeria.

H17: There is a long run equilibrium relationship between United States Output per hour and macroeconomic financial leverage in Nigeria.

H18: There is a long run equilibrium relationship between the United States Financial Soundness Indicators and macroeconomic financial leverage in Nigeria.

## 4.5 MODELLING SHORT RUN DYNAMICS IN FINANCE

The vector error correction model reconciles the long run and short run behaviour of the variables. The error correction model is estimated for both the country specific equations and external factors equations.

## 4.5.1 Vector Error Correction Mechanism for Macroeconomic Financial Leverage and Country Specific Financial Soundness Indicators.

The results of the Vector error correction model for the country specific equation is presented in Tables 17, 18, and 19. It indicates that Macroeconomic Financial leverage is a function of lagged values of Nigerian Government Oil revenue, Nigerian Consumer price index, and Nigerian Mineral Production Index. The cointegrating equation known as the error correction term, (-0.055) as the correct sign and is statistically significant. This implies that 5.5% of the discrepancy between the actual or equilibrium value of Macroeconomic Financial Leverage is eliminated or corrected quarterly.

Therefore the following research hypothesis is not rejected at either 1,5 or 10% level of significance;

H19: The macroeconomic financial leverage and country specific financial soundness indicators can be expressed as a an error correction mechanism.

## 4.5.2 Vector Error Correction Mechanism for Macroeconomic Financial Leverage and External Factors Financial Soundness Indicators.

The results of the Vector error correction model for external factors equation is presented in Tables 21, 22 and 23. It indicates that Macroeconomic Financial leverage is a function of lagged values of United States Foreign Reserve Assets, United States Output per hour, United States Imports, United States Gross Domestic Product, the cointegrating equation known as the error correction term, (-0.042) as the correct sign and is statistically significant. This implies that 4.2% of the discrepancy between the actual or equilibrium value of Macroeconomic Financial Leverage is eliminated or corrected quarterly.

Therefore the following research hypothesis is not rejected at either 1,5 or 10% level of significance;

H20: The macroeconomic financial leverage and external factor financial soundness indicators can be expressed as a an error correction mechanism.

## TABLE 13: EQUATION 14

## Error Correction

## D(NGNGDD)

## D(NGNFD)

## D(NGNCPI)

## D(NGNIMP)

## D(NGNIEC)

## D(NGNIDP)

## D(NGNGOILR)

## D(NGNCAPEX)

## D(NGNREAX)

## CointEq1

(0.00937)*

(2.8E-07)***

(9.4E-06)*

(1.2E-06)

(1.3E-06)

(6.1E-07)

(0.00572)*

(2.1E-07)

(6.3E-07)

[-4.49486]

[-2.10124]

[-4.59076]

[-0.68594]

[ 1.61931]

[1.16163]

[ 3.81409]

[ 1.42108]

[ 0.51925]

## D(NGNGDD(-1))

(0.19669)

(5.8E-06)**

(0.00020)**

(2.6E-05)

(2.6E-05)

(1.3E-05)

(0.12018)**

(4.4E-06)

(1.3E-05)

[ 0.37664]

[ 2.52721]

[-2.88797]

[-0.53356]

[-0.39540]

[-0.23782]

[-2.84627]

[ 0.92497]

[-0.23556]

## D(NGNFD(-1))

(7047.85)

(0.20870)

(7.07238)

(0.92037)

(0.94249)

(0.45704)

(4306.50)

(0.15842)

(0.47544)

[-1.36380]

[-1.12166]

[-0.73778]

[ 1.31388]

[ 0.92489]

[ 0.46493]

[ 0.05511]

[-1.29433]

[-0.89478]

## D(NGNCPI(-1))

(164.895)***

(0.00488)***

(0.16547)***

(0.02153)

(0.02205)

(0.01069)

(100.757)*

(0.00371)

(0.01112)

[-2.19330]

[-2.04329]

[ 2.26096]

[-0.34374]

[ 1.35003]

[0.78272]

[ 4.61277]

[ 0.64998]

[ 0.95557]

## D(NGNIMP(-1))

(1583.63)

(0.04689)

(1.58914)

(0.20680)

(0.21177)***

(0.10270)

(967.658)

(0.03560)

(0.10683)

[ 0.36967]

[ 0.01531]

[ 0.36536]

[-1.46338]

[ 1.93555]

[0.13548]

[-0.20231]

[-0.71391]

[ 0.57707]

## D(NGNIEC(-1))

(1533.26)

(0.04540)

(1.53859)

(0.20023)***

(0.20504)

(0.09943)

(936.878)

(0.03446)

(0.10343)

[-0.99171]

[ 0.20470]

[ 0.12998]

[-2.36197]

[ 1.69117]

[ 0.70553]

[ 1.15071]

[ 1.18635]

[ 1.26439]

## D(NGNIDP(-1))

(3371.92)

(0.09985)

(3.38366)

(0.44033)

(0.45092)

(0.21866)

(2060.37)

(0.07579)

(0.22747)

[ 0.12559]

[ 1.25025]

[-0.27697]

[ 1.41494]

[ 0.00389]

[-0.70578]

[-0.73195]

[ 0.32196]

[-1.49168]

## D(NGNGOILR(-1))

(0.33049)**

(9.8E-06)

(0.00033)*

(4.3E-05)

(4.4E-05)

(2.1E-05)

(0.20194)

(7.4E-06)

(2.2E-05)***

[-2.89832]

[-1.48494]

[-3.13422]

[-0.26442]

[ 1.89461]

[1.24935]

[ 1.15485]

[ 0.88803]

[-2.28598]

## D(NGNCAPEX(-1))

(7827.09)

(0.23177)

(7.85434)

(1.02213)

(1.04669)

(0.50757)

(4782.65)

(0.17594)*

(0.52801)

[-0.30552]

[-0.37273]

[ 0.30482]

[-0.26659]

[-0.19957]

[-0.37073]

[ 0.10478]

[-2.49639]

[ 0.75967]

## D(NGNREAX(-1))

(2297.27)

(0.06802)

(2.30526)**

(0.30000)

(0.30721)

(0.14897)

(1403.72)

(0.05164)

(0.15497)

[-0.40900]

[-0.51584]

[ 2.80131]

[ 0.72165]

[ 0.61845]

[ 0.71902]

[ 1.83078]

[ 0.36441]

[ 0.46469]

## CONSTANT

(49127.1)*

(1.45471)

(49.2981)*

(6.41543)

(6.56961)

(3.18581)

(30018.5)**

(1.10427)

(3.31408)

[ 3.73076]

[ 1.32397]

[ 4.02628]

[ 0.66660]

[-1.07099]

[0.14596]

[-3.03786]

[-0.99477]

[ 0.54697]

## R-squared

0.535012

0.363303

0.829505

0.260085

0.261401

0.103353

0.529743

0.317164

0.393625

## Adj. R-squared

0.362794

0.127489

0.766359

-0.013957

-0.012155

-0.228738

0.355574

0.064262

0.169041

## F-statistic

3.106600*

1.540636***

13.13628*

0.949070

0.955568

0.311220

3.041543*

1.254099

1.752687

## *, * *, *** represents significance level at 1,5 and 10% respectively. Standard errors in ( ), t-statistics in [ ]. CointEqu1 is denoted as the error correction term.

## TABLE 14: EQUATION 15

## Error Correction

## D(NGNGED)

## D(NGNFD)

## D(NGNCPI)

## D(NGNREAX)

## D(NGNCAPEX)

## D(NGNGOILR)

## D(NGNIMP)

## D(NGNIEC)

## D(NGNIDP)

## CointEq1

(0.05089)

(2.9E-07)

(1.0E-05)*

(5.8E-07)

(2.1E-07)

(0.00638)**

(1.2E-06)

(1.1E-06)**

(5.7E-07)

[-0.59389]

[-1.98390]

[-4.14789]

[ 0.23989]

[ 1.30947]

[ 2.56851]

[-0.68949]

[ 2.53877]

[ 1.63328]

## D(NGNGED(-1))

(0.26546)

(1.5E-06)

(5.3E-05)

(3.0E-06)***

(1.1E-06)

(0.03329)

(6.2E-06)

(5.8E-06)***

(3.0E-06)

[ 0.21833]

[ 0.52816]

[ 0.23902]

[-1.77645]

[ 0.10000]

[-1.52180]

[-0.33524]

[ 1.91283]

[ 1.12401]

## D(NGNFD(-1))

(36469.2)

(0.21038)

(7.29120)

(0.41554)

(0.14904)

(4573.80)

(0.85274)

(0.79418)

(0.40761)

[ 0.52496]

[ 0.11680]

[-1.60234]

[-1.44392]

[-1.02861]

[-1.46026]

[ 1.16419]

[ 1.09010]

[ 0.51236]

## D(NGNCPI(-1))

(840.866)

(0.00485)

(0.16811)

(0.00958)

(0.00344)

(105.458)***

(0.01966)

(0.01831)**

(0.00940)

[-1.18880]

[-0.96350]

[ 1.02181]

[ 0.61168]

[ 1.05547]

[ 2.49407]

[-0.65005]

[ 2.17035]

[ 1.19805]

## D(NGNREAX(-1))

(15274.2)

(0.08811)

(3.05373)

(0.17404)

(0.06242)

(1915.62)***

(0.35715)

(0.33262)

(0.17072)

[ 0.24767]

[-0.45627]

[ 1.46544]

[ 1.38441]

[ 0.42159]

[ 1.96258]

[ 0.69390]

[-0.46225]

[ 0.02705]

## D(NGNCAPEX(-1))

(43770.6)

(0.25250)

(8.75097)

(0.49874)

(0.17888)***

(5489.51)

(1.02347)

(0.95318)

(0.48922)

[ 0.34092]

[-0.19464]

[ 0.01277]

[ 0.83161]

[-2.39161]

[-0.14848]

[-0.29715]

[-0.22070]

[-0.38729]

## D(NGNGOILR(-1))

(2.29326)

(1.3E-05)

(0.00046)***

(2.6E-05)**

(9.4E-06)

(0.28761)

(5.4E-05)

(5.0E-05)*

(2.6E-05)***

[ 0.37761]

[-1.17415]

[-2.11336]

[-2.91422]

[ 0.67499]

[ 0.01586]

[-0.35215]

[ 3.13071]

[ 1.89378]

## D(NGNIMP(-1))

(8723.40)

(0.05032)

(1.74405)

(0.09940)

(0.03565)

(1094.05)

(0.20397)

(0.18997)***

(0.09750)

[ 0.30371]

[ 0.57029]

[-0.12045]

[ 0.42243]

[-0.53445]

[-0.84552]

[-1.61139]

[ 2.17478]

[ 0.15262]

## D(NGNIEC(-1))

(9197.65)

(0.05306)

(1.83887)

(0.10480)

(0.03759)

(1153.53)

(0.21506)**

(0.20030)**

(0.10280)

[ 0.25637]

[ 0.07237]

[-0.36640]

[ 0.92651]

[ 1.22341]

[ 0.65292]

[-2.32648]

[ 2.51610]

[ 1.15097]

## D(NGNIDP(-1))

(19272.4)

(0.11118)

(3.85309)

(0.21960)

(0.07876)

(2417.06)

(0.45064)

(0.41969)

(0.21541)

[-0.83349]

[ 1.04594]

[ 0.03449]

[-1.28822]

[ 0.22111]

[-0.27795]

[ 1.47279]

[-0.44687]

[-0.98342]

## CONSTANT

(274897.)

(1.58582)

(54.9596)*

(3.13228)

(1.12341)

(34476.4)***

(6.42779)

(5.98638)

(3.07251)

[ 0.97603]

[ 1.01452]

[ 3.92834]

[ 0.82261]

[-1.02259]

[-1.87715]

[ 0.73053]

[-1.66928]

[-0.12781]

## R-squared

0.152742

0.240032

0.787163

0.455942

0.290174

0.376970

0.253960

0.384019

0.162325

## Adj. R-squared

-0.161057

-0.041437

0.708335

0.254439

0.027276

0.146218

-0.022351

0.155877

-0.147924

## F-statistic

0.486752

0.852782

9.985779*

2.262705**

1.103750

1.633659

0.919108

1.683249

0.523209

## *, * *, *** represents significance level at 1,5 and 10% respectively. Standard errors in ( ), t-statistics in [ ]. CointEqu1 is denoted as the error correction term

## TABLE 15: EQUATION 16

## Error Correction

## D(NGNTBILL)

## D(NGNFD)

## D(NGNCAPEX)

## D(NGNREAX)

## D(NGNCPI)

## D(NGNGOILR)

## D(NGNIMP)

## D(NGNIDP)

## D(NGNIEC)

## CointEq1

(0.03914)

(5.64394)***

(3.95068)

(11.6107)

(156.819)*

(120297.)**

(22.8973)

(11.4330)

(24.3797)

[-0.54004]

[-1.72971]

[ 1.20577]

[ 0.76871]

[-5.80137]

[ 3.23470]

[-0.46827]

[ 0.99882]

[ 0.50250]

## D(NGNTBILL(-1))

(0.16157)

(23.2982)

(16.3085)

(47.9291)

(647.349)*

(496589.)

(94.5204)

(47.1956)

(100.640)

[-1.56953]

[ 1.43932]

[-1.11970]

[-0.84816]

[-3.00310]

[-1.45519]

[-0.65745]

[-0.15553]

[ 0.44870]

## D(NGNFD(-1))

(0.00151)

(0.21806)

(0.15264)

(0.44860)

(6.05895)*

(4647.88)

(0.88468)

(0.44173)

(0.94195)

[-0.72347]

[ 0.05320]

[-1.06979]

[-1.15799]

[-3.95537]

[-0.94053]

[ 0.89412]

[ 0.48625]

[ 0.93791]

## D(NGNCAPEX(-1))

(0.00173)

(0.24898)

(0.17428)**

(0.51220)

(6.91802)

(5306.89)

(1.01011)

(0.50436)

(1.07550)

[ 0.42380]

[-0.03025]

[-2.52910]

[ 0.77635]

[ 0.56433]

[-0.35727]

[-0.21128]

[-0.49645]

[-0.46936]

## D(NGNREAX(-1))

(0.00052)

(0.07570)

(0.05299)

(0.15573)

(2.10332)*

(1613.48)***

(0.30711)

(0.15334)

(0.32699)

[-0.98114]

[-0.68354]

[ 0.93333]

[ 0.72356]

[ 2.15495]

[ 2.04315]

[ 0.66787]

[ 0.86654]

[ 0.52900]

## D(NGNCPI(-1))

(3.3E-05)

(0.00473)

(0.00331)

(0.00973)

(0.13144)

(100.828)**

(0.01919)

(0.00958)

(0.02043)

[-1.50113]

[-0.65725]

[ 0.87081]

[ 1.09207]

[ 0.59212]

[ 3.05946]

[-0.50445]

[ 0.59263]

[ 0.40440]

## D(NGNGOILR(-1))

(8.9E-08)

(1.3E-05)

(8.9E-06)

(2.6E-05)

(0.00035)*

(0.27225)

(5.2E-05)

(2.6E-05)

(5.5E-05)

[-0.41836]

[-1.39402]

[ 0.62371]

[-1.74188]

[-4.72772]

[ 1.48215]

[-0.24915]

[ 1.16157]

[ 1.12750]

## D(NGNIMP(-1))

(0.00035)**

(0.05080)

(0.03556)

(0.10451)

(1.41150)

(1082.78)

(0.20610)

(0.10291)

(0.21944)***

[ 3.12936]

[ 0.24555]

[-0.36072]

[ 0.63133]

[-1.51815]

[-0.17217]

[-1.70938]

[ 0.26916]

[ 1.93033]

## D(NGNIDP(-1))

(0.00075)

(0.10764)

(0.07535)

(0.22144)

(2.99080)

(2294.27)

(0.43669)

(0.21805)

(0.46496)

[-1.28717]

[ 0.98602]

[ 0.31215]

[-1.60654]

[-1.32973]

[-0.26930]

[ 1.22701]

[-0.54229]

[ 0.41366]

## D(NGNIEC(-1))

(0.00035)

(0.05088)

(0.03561)

(0.10467)

(1.41369)

(1084.45)

(0.20641)***

(0.10307)

(0.21978)

[ 1.12715]

[ 0.00628]

[ 1.33412]

[ 1.52844]

[ 0.37858]

[ 1.19807]

[-2.08014]

[ 0.65191]

[ 1.04418]

## CONSTANT

(0.01159)

(1.67173)

(1.17019)

(3.43909)

(46.4497)*

(35632.0)***

(6.78219)

(3.38646)

(7.22127)

[ 1.23176]

[ 0.81658]

[-0.90645]

[ 0.40051]

[ 5.71877]

[-2.34967]

[ 0.63309]

[ 0.19275]

[-0.33377]

## R-squared

0.454713

0.245381

0.311832

0.413971

0.864159

0.405360

0.257860

0.090740

0.199110

## Adj. R-squared

0.252754

-0.034108

0.056955

0.196923

0.813847

0.185122

-0.017007

-0.246023

-0.097515

## F-statistic

2.251519***

0.877963

1.223459

1.907282

17.17614*

1.840559

0.938128

0.269448

0.671251

## *, * *, *** represents significance level at 1,5 and 10% respectively. Standard errors in ( ), t-statistics in [ ]. CointEqu1 is denoted as the error correction term.

## TABLE 16: EQUATION 17

## Error Correction

## D(NGNGDD)

## D(USCPI)

## D(USFORRES)

## D(USOUTHR)

## D(USIMPORTS)

## D(USGDP)

## D(USCAPEX)

## D(USIDPROD)

## D(USGCINV)

## CointEq1

(0.21074)*

(3.6E-06)

(0.01556)

(2.6E-06)

(0.00012)**

(0.00022)**

(6.4E-05)

(4.4E-06)*

(3.7E-05)*

[-7.23838]

[-1.32068]

[ 0.19839]

[-0.85929]

[-2.57866]

[-3.97030]

[-0.25774]

[-2.39124]

[-6.03095]

## D(NGNGDD(-1))

(0.19247)

(3.3E-06)***

(0.01421)

(2.4E-06)

(0.00011)

(0.00020)***

(5.8E-05)

(4.0E-06)

(3.4E-05)**

[ 1.73930]

[-1.90456]

[-0.57979]

[ 1.39474]

[ 1.26299]

[ 2.05896]

[ 0.82908]**

[ 0.76879]

[ 3.21322]

## D(USCPI(-1))

(9705.44)

(0.16538)***

(716.466)

(0.12040)**

(5.36931)

(9.94171)**

(2.94240)

(0.20250)***

(1.69197)***

[-0.02897]

[ 2.22879]

[ 0.14529]

[-3.34677]

[-0.16937]

[-2.98064]

[-1.35599]

[-2.27536]

[ 2.06537]

## D(USFORRES(-1))

(2.81146)**

(4.8E-05)

(0.20755)

(3.5E-05)

(0.00156)

(0.00288)

(0.00085)

(5.9E-05)

(0.00049)*

[ 3.85396]

[ 1.56907]

[ 0.15681]

[ 0.87475]

[ 0.62567]

[ 1.22521]

[-0.19602]

[ 0.29946]

[ 4.12028]

## D(USOUTHR(-1))

(15641.1)***

(0.26652)

(1154.64)

(0.19404)

(8.65308)

(16.0219)

(4.74191)***

(0.32634)

(2.72675)***

[-2.29508]

[-1.68955]

[-0.60681]

[ 0.09703]

[-0.35694]

[-1.32240]

[ 2.00425]

[-1.15442]

[-1.88615]

## D(USIMPORTS(-1))

(630.510)*

(0.01074)***

(46.5449)

(0.00782)

(0.34882)***

(0.64586)

(0.19115)

(0.01316)***

(0.10992)**

[-3.52857]

[-2.05795]

[-1.24732]

[ 1.63852]

[-2.38750]

[-1.50270]

[-0.24420]

[-1.76210]

[-3.06189]

## D(USGDP(-1))

(304.159)***

(0.00518)

(22.4534)

(0.00377)

(0.16827)

(0.31156)

(0.09221)

(0.00635)

(0.05302)*

[ 2.29463]

[ 1.49263]

[ 0.91223]

[-1.20594]

[ 0.61298]

[-0.50672]

[ 0.13542]

[-0.25254]

[ 4.55101]

## D(USCAPEX(-1))

(746.196)

(0.01271)

(55.0850)

(0.00926)

(0.41282)*

(0.76436)*

(0.22622)**

(0.01557)**

(0.13009)**

[ 1.45516]

[ 1.07950]

[-0.80954]

[ 0.50983]

[ 4.28323]

[ 4.34464]

[ 3.09354]

[ 3.24649]

[ 3.16671]

## D(USIDPROD(-1))

(18370.9)

(0.31304)

(1356.16)

(0.22790)

(10.1633)

(18.8181)

(5.56951)

(0.38330)

(3.20265)**

[-0.85002]

[ 0.08879]

[ 0.32230]

[-0.93730]

[ 0.21574]

[ 0.70432]

[-1.42007]

[ 0.61833]

[-3.34631]

## D(USGCINV(-1))

(860.859)

(0.01467)

(63.5496)

(0.01068)

(0.47625)

(0.88182)

(0.26099)

(0.01796)

(0.15008)

[-0.60443]

[ 1.04344]

[ 0.62009]

[ 0.35237]

[-0.99123]

[-1.54774]

[-2.05701]

[-1.24997]

[ 0.49222]

## CONSTANT

(73824.6)

(1.25795)

(5449.81)

(0.91584)*

(40.8418)

(75.6218)*

(22.3814)***

(1.54030)*

(12.8700)

[ 1.18931]

[ 1.61843]

[-0.19913]

[ 4.44790]

[ 1.18878]

[ 6.01524]

[ 2.19156]

[ 3.57188]

[ 1.17733]

## R-squared

0.776708

0.481693

0.206009

0.564619

0.612143

0.730819

0.639627

0.611897

0.866877

## Adj. R-squared

0.694007

0.289727

-0.088061

0.403367

0.468493

0.631123

0.506155

0.468155

0.817573

## F-statistic

9.391766*

2.509267**

0.700543

3.501468**

4.261337*

7.330432*

4.792233*

4.256920*

17.58203*

## *, * *, *** represents significance level at 1,5 and 10% respectively. Standard errors in ( ), t-statistics in [ ]. CointEqu1 is denoted as the error correction term

## TABLE 17: EQUATION 18

## Error Correction

## D(NGNGED)

## D(USIMPORTS)

## D(USCPI)

## D(USOUTHR)

## D(USIDPROD)

## D(USGDP)

## D(USGCINV)

## D(USFORRES)

## D(USCAPEX)

## CointEq1

(0.24897)

(1.4E-05)*

(6.2E-07)

(4.0E-07)

(5.4E-07)*

(3.2E-05)*

(7.5E-06)**

(0.00207)*

(9.9E-06)

[ 0.35206]

[-4.98369]

[-1.69199]

[-0.51047]

[-4.43253]

[-3.85855]

[-2.70009]

[ 3.12428]

[ 0.89490]

## D(NGNGED(-1))

(0.29786)

(1.7E-05)*

(7.5E-07)

(4.8E-07)

(6.4E-07)

(3.9E-05)

(9.0E-06)

(0.00248)***

(1.2E-05)

[-0.16243]

[ 2.43279]

[ 1.01391]

[ 1.30788]

[ 1.52812]

[ 1.60466]

[ 1.06610]

[-2.00864]

[-0.74187]

## D(USIMPORTS(-1))

(4999.63)

(0.28301)*

(0.01252)***

(0.00810)

(0.01076)**

(0.65247)

(0.15125)

(41.6131)

(0.19832)

[-0.14566]

[-4.17562]

[-2.02864]

[ 1.64625]

[-3.11275]

[-1.48239]

[-1.17237]

[ 0.58641]

[ 0.49405]

## D(USCPI(-1))

(69286.4)

(3.92207)

(0.17350)

(0.11229)**

(0.14909)*

(9.04218)*

(2.09609)

(576.687)

(2.74833)

[-0.37191]

[-1.06992]

[ 0.90053]

[-3.35293]

[-4.06331]

[-4.11855]

[ 0.85833]

[ 0.18571]

[-1.14785]

## D(USOUTHR(-1))

(117631.)

(6.65870)

(0.29456)***

(0.19064)

(0.25312)

(15.3514)

(3.55864)

(979.071)

(4.66599)***

[ 0.64044]

[ 0.06793]

[-2.09259]

[ 0.15715]

[-0.94138]

[-0.64113]

[-0.65545]

[-0.81709]

[ 2.36759]

## D(USIDPROD(-1))

(135844.)

(7.68968)

(0.34017)

(0.22016)

(0.29231)

(17.7282)

(4.10964)

(1130.66)

(5.38843)

[-0.94134]

[ 1.00151]

[-0.05730]

[-0.66186]

[ 1.40546]

[ 1.64802]

[-1.60582]

[ 0.24010]

[-1.28957]

## D(USGDP(-1))

(2260.17)

(0.12794)

(0.00566)

(0.00366)

(0.00486)

(0.29496)

(0.06838)**

(18.8120)

(0.08965)

[ 1.45276]

[ 0.07109]

[ 0.72735]

[-1.09380]

[-1.19051]

[-1.44195]

[ 2.55438]

[ 0.99539]

[ 0.17477]

## D(USGCINV(-1))

(5038.82)

(0.28523)

(0.01262)

(0.00817)

(0.01084)

(0.65759)

(0.15244)**

(41.9393)

(0.19987)***

[-1.20670]

[-0.83158]

[ 0.18230]

[ 1.64562]

[-1.66217]

[-0.33404]

[ 2.95358]

[ 1.30887]

[-1.85277]

## D(USFORRES(-1))

(22.4405)

(0.00127)

(5.6E-05)

(3.6E-05)

(4.8E-05)

(0.00293)

(0.00068)

(0.18678)

(0.00089)

[ 0.46099]

[-1.89464]

[-0.02179]

[-0.13014]

[-1.75779]

[-1.35543]

[ 0.48047]

[ 1.22352]

[ 0.18739]

## D(USCAPEX(-1))

(4083.48)

(0.23115)*

(0.01023)***

(0.00662)

(0.00879)*

(0.53291)*

(0.12354)

(33.9878)

(0.16198)**

[-1.67266]

[ 5.35888]

[ 2.18810]

[-1.03555]

[ 4.25739]

[ 3.33162]

[-0.02417]

[-0.64533]

[ 3.67134]

## CONSTANT

(527470.)

(29.8583)**

(1.32085)**

(0.85484)*

(1.13500)*

(68.8372)*

(15.9573)

(4390.26)

(20.9228)

[-0.04544]

[ 3.24540]

[ 3.60654]

[ 4.15871]

[ 6.56638]

[ 7.27479]

[ 0.78766]

[-1.50607]

[ 1.29124]

## R-squared

0.246765

0.763577

0.348279

0.567389

0.759662

0.745614

0.766595

0.412336

0.640819

## Adj. R-squared

-0.03221

0.676013

0.106901

0.407162

0.670648

0.651397

0.680148

0.194683

0.507788

## F-statistic

0.884541

8.720220*

1.442877

3.541168**

8.534177*

7.913798*

8.867858*

1.894462

4.817092**

## *, * *, *** represents significance level at 1,5 and 10% respectively Standard errors in ( ), t-statistics in [ ]. CointEqu1 is denoted as the error correction term.

## TABLE 18: EQUATION 19

## Error Correction

## D(NGNTBILL)

## D(USCPI)

## D(USFORRES)

## D(USOUTHR)

## D(USIMPORTS)

## D(USGDP)

## D(USGCINV)

## D(USCAPEX)

## D(USIDPROD)

## CointEq1

(0.01718)

(0.87957)***

(3431.30)

(0.56678)

(25.1530)**

(51.5646)**

(10.0655)**

(13.7445)

(1.01017)***

[ 0.60263]

[ 1.80523]

[-0.39240]

[-1.53366]

[ 2.83272]

[ 2.90617]

[ 3.71435]

[ 1.11625]

[ 1.81262]

## D(NGNTBILL(-1))

(0.18650)

(9.54564)

(37238.6)

(6.15109)

(272.976)

(559.611)

(109.237)

(149.164)

(10.9630)

[-0.76702]

[ 0.10427]

[ 0.17276]

[ 0.24721]

[ 0.71290]

[ 1.73950]

[ 0.72552]

[-0.52323]

[ 0.48145]

## D(USCPI(-1))

(0.00398)

(0.20375)

(794.849)

(0.13129)***

(5.82661)***

(11.9448)*

(2.33163)

(3.18387)

(0.23400)**

[-0.13056]

[-0.08560]

[ 0.18572]

[-2.05818]

[-2.01537]

[-4.53164]

[-1.04136]

[-1.59267]

[-3.37399]

## D(USFORRES(-1))

(9.2E-07)

(4.7E-05)

(0.18407)

(3.0E-05)

(0.00135)

(0.00277)

(0.00054)

(0.00074)

(5.4E-05)

[-0.84925]

[ 0.75937]

[ 0.14126]

[ 0.53109]

[-0.44133]

[-0.50702]

[ 1.54241]

[-0.07498]

[-0.80516]

## D(USOUTHR(-1))

(0.00576)

(0.29472)**

(1149.73)

(0.18991)

(8.42802)

(17.2777)

(3.37264)***

(4.60537)*

(0.33848)

[ 0.80141]

[-2.54819]

[-0.67383]

[ 0.91916]

[-0.79005]

[-1.50627]

[-1.83917]

[ 2.01336]

[-1.42189]

## D(USIMPORTS(-1))

(0.00019)

(0.00979)

(38.1850)

(0.00631)**

(0.27991)

(0.57383)

(0.11201)

(0.15296)

(0.01124)

[-1.31084]

[-1.75268]

[-1.46257]

[ 3.23474]

[-1.71379]

[ 0.59283]

[-0.31852]

[-0.66415]

[-0.74181]

## D(USGDP(-1))

(0.00011)

(0.00570)

(22.2269)

(0.00367)

(0.16293)

(0.33402)

(0.06520)**

(0.08903)

(0.00654)

[ 1.21088]

[ 0.95005]

[ 0.84897]

[-1.40610]

[ 0.53560]

[-0.62868]

[ 3.29918]

[ 0.28950]

[-0.49639]

## D(USGCINV(-1))

(0.00025)

(0.01278)

(49.8504)

(0.00823)***

(0.36543)

(0.74914)

(0.14623)**

(0.19968)***

(0.01468)

[-1.49099]

[ 0.13860]

[ 0.45597]

[ 2.03681]

[-0.36689]

[-0.28315]

[ 2.82741]

[-2.32403]

[-0.75224]

## D(USCAPEX(-1))

(0.00030)

(0.01548)

(60.4038)

(0.00998)

(0.44279)

(0.90773)

(0.17719)**

(0.24196)***

(0.01778)

[-0.86885]

[ 0.24137]

[-0.18918]

[ 0.49432]

[ 0.66668]

[-0.53857]

[-2.76752]

[ 1.80984]

[ 0.60301]

## D(USIDPROD(-1))

(0.00675)

(0.34565)

(1348.42)

(0.22273)

(9.88453)

(20.2636)

(3.95549)***

(5.40126)

(0.39697)

[-0.91063]

[-0.26107]

[ 0.21560]

[-0.50402]

[ 0.34822]

[ 0.82906]

[-2.19313]

[-1.31880]

[ 0.73228]

## CONSTANT

(0.03106)

(1.58986)**

(6202.24)

(1.02449)***

(45.4653)***

(93.2055)*

(18.1939)

(24.8439)

(1.82593)**

[ 0.49692]

[ 3.52413]

[-0.12978]

[ 2.11065]

[ 2.41693]

[ 5.98882]

[ 2.10588]***

[ 2.28414]***

[ 3.92807]

## R-squared

0.281093

0.358326

0.202949

0.577740

0.627472

0.683066

0.793805

0.655845

0.577293

## Adj. R-squared

0.014831

0.120669

-0.092254

0.421347

0.489499

0.565682

0.717436

0.528381

0.420734

## F-statistic

1.055699

1.507746

0.687489

3.694166**

4.547781*

5.819112*

10.39438*

5.145308*

3.687396**

## *, * *, *** represents significance level at 1,5 and 10% respectively Standard errors in ( ), t-statistics in [ ]. CointEqu1 is denoted as the error correction term

## 4.6 MODELLING CAUSALITY IN FINANCE

The main essence of causality is to determine the direction of influence amongst the variables in the study. The Granger Causality tests shows the direction of influence among the variables, in times series analysis, it is assumed that one event causes another event.

## 4.6.1 Granger Causality Tests For Country Specific Financial Soundness Indicators

The results of the Granger Causality tests for the country specific equations is presented in Table 23, they are as follows;

Bilateral Causality; There exists a bilateral or feedback causality between Macroeconomic Financial Leverage and Financial Deepening index, and Government Oil revenue and Consumer Price Index and Industrial production index.

Unilateral Causality; There exists a unilateral causality between Macroeconomic Financial Leverage and Mineral Production index, and Energy Consumption index and Capital expenditure.

Independence; The results of the Granger Causality tests shows an independent relationship between Macroeconomic Financial leverage and Real exchange rate in Nigeria

## Table 23

## PAIRWISE CASUALITY- Country Specific Equations- 7, 8, 9.

## Lags: 3

## Null Hypothesis

## F-Statistic

## Probability

NgnFD does not Granger Cause NgnTbill

NgnTbill does not Granger Cause NgnFD

## 3.4598** 2.3573***

## 0.0285

## 0.0916

NgnFD does not Granger Cause NgnGDD NgnGDD does not Granger Cause NgnFD

## 0.5068 1.7526

## 0.6805

## 0.1775

NgnFD does not Granger Cause NgnGED

NgnGED does not Granger Cause NgnFD

## 6.7685 1.6741

## 0.5207

## 0.1936

NgnGOilr does not Granger Cause NgnTbill NgnTbill does not Granger Cause NgnGOilr

## 0.2730 4.4470*

## 0.8443

## 0.0106

NgnGOilr does not Granger Cause NgnGDD NgnGDD does not Granger Cause NgnGOilr

## 4.1372* 3.6581**

## 0.0144 0.0233

NgnGOilr does not Granger Cause NgnGED NgnGED does not Granger Cause NgnGOilr

## 1.4641 3.3867**

## 0.2441

## 0.0308

NgnCPI does not Granger Cause NgnTbill

NgnTbill does not Granger Cause NgnCPI

## 1.7271 4.6009*

## 0.1824 0.0091

NgnCPI does not Granger Cause NgnGDD NgnGDD does not Granger Cause NgnCPI

## 4.5545* 2.9503**

## 0.0019 0.0485

NgnCPI does not Granger Cause NgnGED

NgnGED does not Granger Cause NgnCPI

## 0.1075 14.523*

## 0.9550

## 0.0000

NgnIMP does not Granger Cause NgnTbill

NgnTbill does not Granger Cause NgnIMP

## 5.1127* 0.9222

## 0.0056 0.4420

NgnIMP does not Granger Cause NgnGDD NgnGDD does not Granger Cause NgnIMP

## 0.1263 1.4818

## 0.9437 0.2294

NgnIMP does not Granger Cause NgnGED NgnGED does not Granger Cause NgnIMP

## 0.1188 0.6468

## 0.9483 0.5912

NgnIDP does not Granger Cause NgnTbill NgnTbill does not Granger Cause NgnIDP

## 4.3324* 7.0902*

## 0.0021 0.0051

NgnIDP does not Granger Cause NgnGDD NgnGDD does not Granger Cause NgnIDP

## 6.1993* 0.6437

## 0.0061

## 0.5929

NgnIDP does not Granger Cause NgnGED NgnGED does not Granger Cause NgnIDP

## 0.3876 0.7138

## 0.7627 0.5514

NgnIEC does not Granger Cause NgnTbill NgnTbill does not Granger Cause NgnIEC

## 0.1297 5.0663*

## 0.9417 0.0773

NgnIEC does not Granger Cause NgnGDD NgnGDD does not Granger Cause NgnIEC

## 0.0798 0.8685

## 0.9704 0.4682

NgnIEC does not Granger Cause NgnGED NgnGED does not Granger Cause NgnIEC

## 6.6147* 1.5586

## 0.0008

## 0.2199

NgnREAX does not Granger Cause NgnTbill NgnTbill does not Granger Cause NgnREAX

## 0.5741 0.2336

## 0.6364 0.8722

NgnREAX does not Granger Cause NgnGDD NgnGDD does not Granger Cause NgnREAX

## 1.1691 0.7623

## 0.3379 0.3241

NgnREAX does not Granger Cause NgnGED NgnGED does not Granger Cause NgnREAX

## 1.0021 0.1358

## 0.4054 0.9379

NgnCAPEX does not Granger Cause NgnTbill NgnTbill does not Granger Cause NgnCAPEX

## 4.0768 * 1.0282

## 0.0021 0.3941

NgnCAPEX does not Granger Cause NgnGDD NgnGDD does not Granger Cause NgnCAPEX

## 0.2410 0.5357

## 0.8671 0.6613

NgnCAPEX does not Granger Cause NgnGED NgnGED does not Granger Cause NgnCAPEX

## 5.2397* 0.4823

## 0.0079 0.6973

## *, * *, *** represents significance level at 1,5 and 10% respectively.

Therefore the following research hypothesises is not rejected at either 1,5 or 10% level of significance;

H21: Nigerian Government Oil revenue granger causes macroeconomic financial leverage in Nigeria

H22: Nigerian Financial deepening granger causes macroeconomic financial leverage in Nigeria.

H23: Nigerian Industrial production granger causes macroeconomic financial leverage in Nigeria.

H24: Nigerian Energy Consumption granger causes macroeconomic financial leverage in Nigeria.

H25: Nigerian Principal Mineral production granger causes macroeconomic financial leverage in Nigeria.

H26: Nigerian Consumer Price Index granger causes macroeconomic financial leverage in Nigeria.

H27: Government capital expenditure granger causes macroeconomic financial leverage in Nigeria.

H29: The Nigerian Financial Soundness Indicators impacts macroeconomic financial leverage in Nigeria.

However, the following hypothesis below is rejected at 1, 5 and 10% level of significance.

H28: The NGN/US real exchange rate granger causes macroeconomic financial leverage in Nigeria.

## 4.6.2 Granger Causality Tests For External Factors Financial Soundness Indicators

The results of the Granger Causality tests for the external factor equations is presented in Table 24, they are as follows;

Bilateral Causality; There exists a bilateral or feedback causality between Macroeconomic Financial Leverage and Foreign Reserve assets and Imports of Goods and Services.

Unilateral Causality; There exists a unilateral causality between Macroeconomic Financial Leverage and between Macroeconomic Financial Leverage and United States Foreign Reserve Assets, and United States Capital Expenditure, and United States Imports, and United States Gross Domestic Product, and United States Government Consumption and Investment, and United States Industrial Production.

Independence; The results of the Granger Causality tests shows independence between Macroeconomic Financial leverage and United States Output per hour.

## Table 24

## PAIRWISE CASUALITY- External Factors Equations 10, 11, 12

## Lags: 10

## Null Hypothesis

## F-Statistic

## Probability

USCPI does not Granger Cause NgnTbill

NgnTbill does not Granger Cause USCPI

## 0.3139 0.3377

## 0.8152

## 0.7981

USCPI does not Granger Cause NgnGDD NgnGDD does not Granger Cause USCPI

## 0.3815 5.1413*

## 0.7670

## 0.0055

USCPI does not Granger Cause NgnGED

NgnGED does not Granger Cause USCPI

## 0.9911 4.1991*

## 0.4103

## 0.0135

USFORRES does not Granger Cause NgnTbill NgnTbill does not Granger Cause USFORRES

## 4.0707* 2.8712**

## 0.0075

## 0.0528

USFORRES does not Granger Cause NgnGDD NgnGDD does not Granger Cause USFORRES

## 4.8002*

## 0.7022

## 0.0076

## 0.5582

USFORRES does not Granger Cause NgnGED NgnGED does not Granger Cause USFORRES

## 0.8193 0.4256

## 0.4934

## 0.7360

USCAPEX does not Granger Cause NgnTbill

NgnTbill does not Granger Cause USCAPEX

## 0.6452 0.7726

## 0.5921 0.5185

USCAPEX does not Granger Cause NgnGDD NgnGDD does not Granger Cause USCAPEX

## 3.8617* 0.6866

## 0.0019 0.5673

USCAPEX does not Granger Cause NgnGED

NgnGED does not Granger Cause USCAPEX

## 0.3825 3.6428**

## 0.7663

## 0.0237

USIDPROD does not Granger Cause NgnTbill

NgnTbill does not Granger Cause USIDPROD

## 0.1786 3.0001**

## 0.9100 0.0460

USIDPROD does not Granger Cause NgnGDD NgnGDD does not Granger Cause USIDPROD

## 0.3646 7.2838*

## 0.7790 0.0008

USIDPROD does not Granger Cause NgnGED NgnGED does not Granger Cause USIDPROD

## 1.0355 1.2644

## 0.3910 0.3043

USGDP does not Granger Cause NgnTbill NgnTbill does not Granger Cause USGDP

## 0.7660 3.0368**

## 0.5220 0.0443

USGDP does not Granger Cause NgnGDD NgnGDD does not Granger Cause USGDP

## 0.2207 6.9055*

## 0.8812

## 0.0011

USGDP does not Granger Cause NgnGED NgnGED does not Granger Cause USGDP

## 0.9857 1.2268

## 0.4127 0.3171

USGCINV does not Granger Cause NgnTbill NgnTbill does not Granger Cause USGCINV

## 0.2540 0.3846

## 0.8578 0.7648

USGCINV does not Granger Cause NgnGDD NgnGDD does not Granger Cause USGCINV

## 0.3365 7.6299*

## 0.7990 0.0006

USGCINV does not Granger Cause NgnGED NgnGED does not Granger Cause USGCINV

## 0.2828 4.0767*

## 0.8137

## 0.0153

USOUTHR does not Granger Cause NgnTbill NgnTbill does not Granger Cause USOUTHR

## 1.2437 0.2627

## 0.3113 0.8517

USOUTHR does not Granger Cause NgnGDD NgnGDD does not Granger Cause USOUTHR

## 0.7489 1.3757

## 0.5315 0.2692

USOUTHR does not Granger Cause NgnGED NgnGED does not Granger Cause USOUTHR

## 1.5138 1.3040

## 0.2311 0.2913

USIMPORTS does not Granger Cause NgnTbill NgnTbill does not Granger Cause USIMPORTS

## 4.8317* 2.8976**

## 0.0070 0.0513

USIMPORTS does not Granger Cause NgnGDD NgnGDD does not Granger Cause USIMPORTS

## 0.4367 6.8298*

## 0.7283 0.0012

USIMPORTS does not Granger Cause NgnGED NgnGED does not Granger Cause USIMPORTS

## 0.8626 4.2406*

## 0.4712 0.0130

## *, * *, *** represents significance level at 1,5 and 10% respectively.

Therefore the following research hypothesises are not rejected at either 1,5 or 10% level of significance;

H30: The United States Gross domestic product granger causes macroeconomic financial leverage in Nigeria.

H31: The Unites States Foreign reserve assets granger causes macroeconomic financial leverage in Nigeria.

H32: The United States Consumer price index granger causes macroeconomic financial leverage in Nigeria.

H33: The United States Government capital expenditure granger causes macroeconomic financial leverage in Nigeria.

H34: The United States Government Consumption and Investment granger causes macroeconomic financial leverage in Nigeria.

H35: The United States Government Imports of goods and services granger causes macroeconomic financial leverage in Nigeria.

H36: The United States Industrial Production granger causes macroeconomic financial leverage in Nigeria.

H38: The United States Financial Soundness Indicators impacts macroeconomic financial leverage in Nigeria.

However, the following hypothesis below is rejected at 1, 5 and 10% level of significance.

H37: The United States Output per Hour granger causes macroeconomic financial leverage in Nigeria.

## 4.7 MODELLING VOLATILITY IN FINANCE

## 4.7.1 Tests For ARCH Effects

Tests for ARCH effects employing lags of 3, 5 ,10 shows evidence of arch effects, in Table 25, therefore we reject the Null hypothesis of no ARCH effects at 1% level of significance. This suggests that the returns on oil price in Nigeria suffer from Heteroscedasticity. Therefore, the GARCH methodology can be applied to modelling oil price volatility in Nigeria. This is consistent with the results of Narayan and Narayan (2007), work on modelling oil price volatility.

The figures 1.9 [82] shows evidence of volatility clustering in the conditional variance graph, the figure 2.0 [83] also shows evidence of volatility clustering for the Pre-Financial Crisis period and the Global Oil Increase period. This means that periods of large swings in volatility is followed by period of large swings and small swings in oil price is followed by periods of small swings in volatility.

For the Global Financial Crisis period, the conditional variance graph and GARCH graph in the Figures 2.1 [84] and 2.2 [85] respectively shows evidence of volatility clustering. For, the total sample period, the Figures 2.3 [86] and 2.4 [87] respectively also show evidence of volatility clustering in the Oil price in Nigeria for the period 30/11/2000 to 07/07/2010.

H39: There is evidence of Autoregressive Conditional Heteroscedasticity (ARCH) effects in Oil price returns in Nigeria.

H40: There is evidence of volatility Clustering in Nigerian Bonny Light Oil Price.

## 4.7.2 Bollerslev-Wooldridge Robust Standard Errors and Variance

To obtain robust inference about the estimated models, the robust standard errors employed in this study are computed by Bollerslev-Wooldridge and as suggested by Engle Granger (1982), Narayan and Narayan (2007), they are employed to obtain robust estimates.

## 4.7.3 The EGARCH Model

The EGARCH model is employed in this research because its measures asymmetric effects and persistency of shocks to oil price volatility.

Following the methodology of Engle (1982), Bollerslev (1986), and Narayan and Narayan (2007) to remedy the presence of Heteroscedasticity, we model oil price return within an GARCH/EGARCH methodology as suggested by Nelson (1991).

## 4.7.3.1 Asymmetric Effects of Oil price Shock

For the total sample, (30/11/2000 to 07/07/2010), γ- which measures asymmetry of shocks is negative (-0.048484) and statistically significant at 1% percent. The negative sign suggests that negative shocks reduce volatility more than positive shocks. This suggests that shocks have asymmetric effects on the volatility of crude oil prices. For the sub-samples, γ- is also negative and statistically significant at 1% , it is -0.048526 for Pre-Financial Crisis & Global Oil Increase Period (30/11/2000 to 30/01/2006) and decreases to -0.050315 for the Financial Crisis & Global Warming Increase Period (01/02/2006 to 07/07/2010). This implies that the role of positive shocks in reducing crude oil price volatility was higher over the pre-financial crisis and global oil increase period than the financial crisis and global warming increase period. This suggests that shocks have asymmetric effects on the volatility of crude oil prices. This result is consistent with empirical studies by Tatom (1988), Mork (1989) Mork, Olsen and Mysen (1994), Ferderer (1996) and Hooker (1996).

Therefore the following research hypothesis is not rejected at either 1,5 or 10% level of significance;

H41: There is evidence of asymmetric effects of shocks to oil price volatility in Nigeria.

## 4.7.3.2 Persistent Effects of Oil Price Shock

For the total sample, (30/11/2000 to 07/07/2010), β – which captures persistence of shocks is positive (0.986290), and statistically significant at 1% percent, the value is also close to 1, suggesting that shocks to crude oil price volatility do not die out rapidly, rather shocks tend to persist. This implies that shocks have permanent effects on crude oil price volatility. For the subsamples also, β - is also positive and statistically significant at 1% percent, it is 0.96776 for the Pre-Financial Crisis & Global Oil Increase Period (30/11/2000 to 30/01/2006) and increases to 0.98982 for the Financial Crisis & Global Warming Increase Period (01/02/2006 to 07/07/2010). This implies that the persistent of shocks increased dramatically from the Pre-Financial Crisis & Global Oil Increase Period (30/11/2000 to 30/01/2006) to the Financial Crisis & Global Warming Increase Period (01/02/2006 to 07/07/2010). The values are also close to 1, suggesting that shocks to crude oil price volatility do not die out rapidly, rather shocks tend to persist t. This also implies that shocks have permanent effects on crude oil price volatility.

H42: There is evidence of persistency effects of shocks to oil price volatility in Nigeria.

## 4.7.3.3 Global Financial Crisis and Oil price Volatility in Nigeria

The Global Financial Crisis has been argued to affect the economy of Nigeria through the volatility of oil price, based on this study evidence is provided that the asymmetrics and persistency of shocks to oil price volatility increases from the pre financial crisis period to the Global financial crisis period. The Figures 2.1 and 2.2 also show evidence of volatility clustering in the Global Financial Crisis period.

H43: The global financial crisis affects the Nigerian economy through the volatility of oil price in Nigeria

## Table 25- EGARCH Model

## MODELLING OIL PRICE VOLATILITY-EGARCH (1,1)

## Parameters Of Mean and Variance Equations

## Full Sample Period 30/11/2000 to 07/07/2010

## Pre-Financial Crisis & Global Oil Increase Period 30/11/2000 to 30/01/2006

## Financial Crisis & Global Warming Increase Period 01/02/2006 to 07/07/2010

## λ

## 0.042480* (0.0062)

## 0.053272 (0.1118)

## 0.038053* (0.0005)

## С

## 0.000418** (0.0221)

## 0.000519** (0.0379)

## 0.000349 (0.1797)

## δ

## 0.941365* (0.0000)

## 0.905002* (0.0000)

## 0.953700* (0.0000)

## ϑ

## 0.007748 (0.71117)

## -0.005759

## (0.8433)

## 0.0023803 (0.4307)

## ω

## -0.180486*

## (0.0003)

## -0.369244**

## (0.0477)

## -0.141363*

## (0.0019)

## α

## 0.067801* (0.0009)

## 0.090922* (0.00554)

## 0.056226* (0.0040)

## γ

## -0.048484*

## -0.048526*

## -0.050315*

## (0.0000)

## (0.00358)

## (0.0001)

## β

## 0.986290*

## 0.96776*

## 0.98982*

## (0.0000)

## (0.0000)

## (0.0000)

## Diagnostic Test: ARCH effects

## Lags

## 3

## 44.48491*

## 16.04908*

## 7.941799*

## (0.0000)

## (0.0000)

## (0.0000)

## 5

## 38.98073*

## 11.49556*

## 7.972328*

## (0.0000)

## (0.0000)

## (0.0000)

## 10

## 25.76900*

## 10.11140*

## 5.681678*

## (0.0000)

## (0.0000)

## (0.0000)

## *, * *, *** represents significance level at 1,5 and 10% respectively. P- values ( ).

## CHAPTER 5

## SUMMARY, CONCLUSION AND RECOMMENDATION

## 5.1 Introduction

This study examines the relationship between macroeconomic financial leverage, global financial crisis and oil price volatility in an emerging market economy, with the case of Nigeria. The previous chapters in this study have examined, the aims and objectives, the research questions and hypotheses, the review of both theoretical and empirical literature was carried out, the research methodology was also discussed in details. The results of the various tests employed to achieve the research objectives was also presented.

The first step in the empirical analysis involved the descriptive statistics, and the serial correlation tests and other stability and residual tests, also testing the time series characteristics of the data series using Augmented Dickey Fuller (ADF), Philips Perrons (PP) and Kwiatkowski-Phillips-Schmidt tests (KPSS) tests for Stationarity, followed by the Engle Granger, Durbin Watson and Johansen tests for cointegration. The Error Correction Mechanism and Granger Causality tests are also conducted.

All the variables employed were characterized by a unit root at level, but the hypothesis of non-stationarity was rejected at first difference. This is consistent with the results of previous empirical studies on time series analysis. There is also evidence of cointegration among the variables which implies a long run equilibrium relationship exists among the variables for the period under study (1970-2009). The Error Correction Mechanism gives evidence that the error correction term, as the correct sign and is statistically significant, this implies that long run equilibrium condition influences the short run dynamics. Therefore, Macroeconomic Financial leverage in Nigeria has an automatic adjustment mechanism and the economy responds to deviations from equilibrium in a balancing manner.

This study also provides evidence that the global financial crisis impacts macroeconomic financial leverage, through financial soundness indicators developed by the International Monetary Fund, World Bank and European Central Bank for the Nigerian (country-specific) economy and the United States (external factors) economy for the period under study. Granger Causality tests represents the direction of influence among the variables in the study, it demonstrates that there exists bilateral and unilateral causality between macroeconomic financial leverage and certain financial soundness indicators.

This study also found evidence of volatility clustering in Nigerian oil price, and shocks to oil price volatility have asymmetric and persistent effects employing the EGARCH technique. By examining, two sub-sample periods and a full sample period the robustness of the results is determined. This study also provides evidence that the persistency of shocks to oil price volatility increases from the pre-financial crisis and global oil increase (30/11/2000 to 30/01/2006) to the financial crisis and global warming increase period (01/02/2006 to 07/07/2010). Thus, this study confirms that the global financial crisis affects the Nigerian economy, through the volatility of oil price for the period under study. It also confirms the IMF and World Bank forecasts [88] of the persistency of shocks beginning from the Global Oil increase period to the Global Financial crisis period.

## 5.2 FINDINGS

## Unit roots/Firsts Differences

All Financial Soundness indicators FSIs in this study for both the country specific and external factors contain a unit root, however taking first differences they become stationary employing the Augmented Dickey Fuller, Philips Perron, and Kwiatkowski-Phillips-Schmidt test for unit roots.

## Long Run Relationships/Cointegration

There is a long run equilibrium relationship between macroeconomic financial leverage in Nigeria and the Country- specific Financial Soundness indicators FSIs employing the Engle Granger, Durbin Watson and Johansen Cointegration tests. There is a long run equilibrium relationship between macroeconomic financial leverage in Nigeria and the External Factors Financial Soundness indicators FSIs employing the Engle Granger, Durbin Watson and Johansen Cointegration tests.

## Short Run Dynamics/Error Correction Model

The 5.5% discrepancy between the actual or equilibrium value of Macroeconomic Financial Leverage is eliminated or corrected quarterly based on the Vector Error Correction Methodology for the Country Specific Financial Soundness Indicators FSIs. The 4.2% discrepancy between the actual or equilibrium value of Macroeconomic Financial Leverage is eliminated or corrected quarterly based on the Vector Error Correction Methodology for the External Factors Financial Soundness Indicators FSIs.

## Causality/Direction of Influence

There exists a bilateral or feedback causality between Macroeconomic Financial Leverage and Nigerian Financial Deepening index, and Nigerian Government Oil revenue and Nigeria Consumer Price Index and Nigerian Industrial production index, United States Foreign Reserve assets and United States Imports of Goods and Services. There exists a unilateral causality between Macroeconomic Financial Leverage and Nigerian Mineral Production index, and Nigerian Energy Consumption index and Nigerian Capital expenditure, United States Capital Expenditure, and United States Gross Domestic Product, and United States Government Consumption and Investment, and United States Industrial Production. There exists independence of causality between Macroeconomic Financial leverage and Nigerian Real exchange rate and United States Output per hour.

## Volatility Clustering/Autoregressive Conditional Heteroscedasticity(ARCH)

There exists volatility clustering in Bonny Light Nigerian oil prices, which means periods of large swings are followed by large swings and small swings are followed by small swings in oil price volatility employing the Autoregressive Conditional Heteroscedascity Methodology. It also means that the variance of the times series varies over time.

## Asymmetrics of Oil price Shocks/ Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH)

There exists consistent asymmetric effects to oil price volatility in Nigeria across the sub samples and the total sample period in this study, employing the Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH) methodology.

## Persistency of Oil price Shocks/ Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH).

There exists consistent persistent effects to oil price volatility in Nigeria across the sub samples and the total sample period in this study, employing the Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH) methodology.

## Global Financial Crisis and the Nigerian Economy.

The role of positive shocks in reducing oil price volatility was higher than negative shocks over the Pre-Financial Crisis period than the Global Financial Crisis period, demonstrating evidence of asymmetric effects. The persistent of shocks increased dramatically from the Pre-Financial Crisis to the Global Financial Crisis. This provides evidence that the Global Financial Crisis affects Oil price volatility in Nigeria.

## Global Oil Increase and the Nigerian Economy.

The role of positive shocks in reducing oil price volatility was higher than negative shocks over the Global Oil Increase period demonstrating evidence of asymmetric effects. The persistent of shocks also increased dramatically from the Global oil increase to the Global Financial Crisis. This provides evidence that the Global Oil Increase affects Oil price volatility in Nigeria.

## Global Warming Increase and the Nigerian Economy.

The role of positive shocks in reducing oil price volatility was high than negative shocks over Global Warming Increase period, demonstrating evidence of asymmetric effects. The persistent of shocks increased dramatically from the Pre-Financial Crisis to the Global Warming Increase period. This provides evidence that the Global Warming Increase affects Oil price volatility in Nigeria.

## 5.2.1 POLICY IMPLICATIONS

The United States being the major importer of the Nigerian oil, this study provides evidence that the economic and financial status of the US economy, affects the Nigerian economy. The impact of the financial crisis on the volatility of oil prices, further confirms the vulnerability of the Nigerian economy. Slow growth in the United States translates to their demand for imports, with Nigeria relying heavily on the US market as destinations for their exports, will be adversely affected. The financial crisis also affects macroeconomic financial leverage, through government oil revenue, therefore making investment and project planning vague affecting financial and macroeconomic policies.

The behaviour of oil prices tend to change over short periods of time, therefore investment decisions need to consider the nature of oil price behaviour. Persistent oil price shocks could have severe macroeconomic implications, therefore inducing several challenges for both fiscal and monetary policy in oil exporting as well importing countries. Positive shocks have a different effect on volatility than negative shocks, this finding is consistent with studies that have demonstrated rising oil prices have negative effect on economic growth and development.

Oil price volatility translates to Oil revenue. Managing Oil revenue volatility, is a key challenge for policymakers in Nigeria, because oil revenue is used to accumulate foreign exchange reserves to lower vulnerability to financial crisis and debt overhang problems.

## 5.3 RECOMMENDATIONS

Given the importance of oil in the Nigerian economy, and providing evidence that global financial crisis influences oil price volatility, it is recommended that greater diversification of the economy, through judicious investment in other productive sectors of the economy, such as Agriculture and Manufacturing should be enhanced. The dependence of the economy on one single export sector results in exposure to excessive volatility. Therefore, it is also recommended that diversification of the export market for oil resource will reduce the vulnerability of the Nigerian economy to both internal and external shocks.

Government should continue pursuing policies to stabilize domestic oil price through subsidization and thus help enhance investment, employment and growth. It is also important for financial market participants to understand the volatility transmission mechanism over time and across series in order to make optimal portfolio allocation decisions. The financial crisis provides an opportunity to plan and create new financial architecture that will better meet the needs for growth and stability in the Nigerian economy and address the vulnerability of the financial systems to instability and crisis.

## 5.4 CONCLUSION

This study contributes to literature in various ways, by focusing on macroeconomic financial leverage in Nigeria and examining its relationship with the global financial crisis and oil price volatility, employing the Financial Soundness Indicators as developed by the International Monetary Fund, World Bank and European Central Bank in 1998 . By employing current data from the period 1970 to 2010, this study employs the Cointegration technique and Error Correction Methodology to model the short run and long run relationships among the indicators. Granger Causality Tests, is also employed to model causality and direction of influence among the Financial Soundness Indicators and Macroeconomic Financial Leverage.

This study also contributes by examining the relationship between the Financial Soundness Indicators of the United States economy and Macroeconomic Financial leverage in Nigeria, by also employing the Cointegration technique, Error Correction Methodology and Granger Causality Tests.

This study further contributes to literature by modelling oil price volatility in Nigeria, by employing current daily observations from the period 30/11/2000 to 07/07/2010. The Tests for ARCH [89] effects presents that Nigerian oil price suffers from Heteroscedasticity, the EGARCH [90] model hence was applied for modelling oil price volatility inorder to examine the asymmetrics and persistency of shocks to volatility.

This study also contributes by employing distinctive and relevant sample periods, the 1970s, represents the end of the Nigerian Biafran War, the 1980s- 1990s represents the Financial Liberalisation period, 2000s represents the Global Oil increase period, Global Warming increase period and the Global Financial crisis period.

## 5.4.1 LIMITATIONS OF THE STUDY

The major limitations of this study is time and resource constraint.

## 5.4.2 SUGGESTIONS FOR FURTHER RESEARCH

This study has attempted to examine the relationship between Macroeconomic financial leverage, Global Financial Crisis and Oil price volatility in Nigeria based on the period under review. In as much as, this study significantly adds to knowledge, in terms of the primary direction of the study, the methodology of research and sample employed I would like to make a few suggestions for future and further studies in this respect;

The researcher can examine the impact of Corruption on Macroeconomic Financial Leverage in Nigeria.

The researcher can compare the empirical results of this study to examining the same area in an another emerging market economy dependent on oil revenue like Algeria and Venezuela or other Organization of Petroleum Exporting Countries (OPEC) member countries.

The use of more Financial Soundness Indicators in addition to the ones used in this study.

## APPENDIX A: Figures in this Study

## Figure 1.1- Nigerian Oil Price: Bonny Light, Dates of Substantial Oil Change.

price

global oil increase

pipeline vandalism

militant takeover

shutdown of refineries

oil workers strike

0

20

40

60

80

100

1995

2000

2005

2010

year

## Figure 1.2- Indicator of Oil Dependence In Nigeria.

## Figure 1.3- Stationarity Residual Graph for Equation 5.

## Figure 1.4- Stationarity Residual Graph for Equation 6.

## Figure 1.5- Stationarity Residual Graph for Equation 7.

## Figure 1.6- Stationarity Residual Graph for Equation 8.

## Figure 1.7- Stationarity Residual Graph for Equation 9.

## Figure 1.8- Stationarity Residual Graph for Equation 10.

## Figure 1.9- Conditional Variance Graph for Pre- Financial Crisis Period and Global Oil Increase Period.

## Figure 2.0- GARCH Graph for Pre- Financial Crisis Period and Global Oil Increase Period .

## Figure 2.1 - Conditional Variance Graph for Global Financial Crisis Period and Global Warming Increase Period.

## Figure 2.2 – GARCH Graph for Global Financial Crisis Period and Global Warming Increase Period.

## Figure 2.3 - Conditional Variance Graph for Total Sample Period.

## Figure 2.4- GARCH Graph for Total Sample Period

## APPENDIX B: Definitions of Variables in this Study;

## Table 26

## Country-specific financial soundness indicators;

## VARIABLES

## FULL MEANING

## DEFINITION

NGNGDD

Nigerian Government Domestic Debt.

A combination of Treasury Certificates and Development Stocks.

NGNGED

Nigerian Government External Debt.

A combination of debt to london paris club, African Development, World Bank, Multilateral debt.

NGNTBILL

Nigerian Interest Rate on Treasury Bills.

The interest rate issued by the Government on treasury bills issued.

NGNFD

Nigerian Financial Deepening Index.

The ratio of money supply to Gross Domestic Product.

NGNCPI

Nigerian Consumer Price Index.

The index prices of all items in all sectors of the economy.

NGNREAX

Nigerian Real Exchange Rate

The average of official exchange rate of the Naira vis a vis the United States Dollar.

NGNIDP

Nigerian Industrial Production index.

A combination of manufacturing, mining, and electricity production.

NGNIMP

Nigerian Mineral Production index.

A combination of coal, marble, casserite, limestone, and Columbite production.

NGNGOILR

Nigerian Government Oil Revenue.

A combination of coal, marble, casserite, limestone, and Columbite production.

NGNCAPEX

Nigerian Government Capital Expenditure.

The expenditure on major Government Construction Projects.

NGNIEC

Nigerian Energy Consumption Index.

The consumption of energy in Nigeria.

## External Factor financial soundness indicators;

## Table 27

## VARIABLES

## FULL MEANING

## DEFINITION

USOUTHR

United States Output per Hour.

The measure of changes in

the relationship between output and the hours expended in

producing that output.

USCAPEX

United States Government Capital Expenditure.

The expenditure on major Government Projects.

USIMPORTS

United States Imports of Goods and Services.

The total of imports of Goods and services.

USGDP

United States Gross Domestic Product.

It is the total expenditures for all goods and services produced, the value added to input materials of production plus taxes, minus government subsidies.

USCPI

United States Consumer Price Index

The index prices of all items in all sectors of the economy.

USGCINV

United States Government Consumption and Investment.

This is the total of all consumption and investment by the Government in the United States.

USFORRES

United States Foreign Reserve Assets.

The foreign currency deposits, Special Drawing rights, Gold stock and bonds held by federal reserve board.

USIDPROD

United States Industrial production Index.

A combination of manufacturing, mining, and utilities production.

## APPENDIX C: Equations in this Study

## Equation 1- Box- Jenkins Methodology.

## Equation 2- Assumption of No Serial Correlation.

## Equation 3- Violation of Serial Correlation.

## Equation 4- Unit Root Specification.

## Equation 5- Augmented Dickey Fuller.

## Equation 6- General Multiple Regression.

## Equation 7- Country Specific Equation with NGNGDD.

## Equation 8- Country Specific Equation with NGNGED.

## Equation 9- Country Specific Equation with NGNTBILL.

## Equation 10- External Factor Equation with NGNGDD.

## Equation 11- External Factor Equation with NGNGED.

## Equation 12- External Factor Equation with NGNTBILL.

## Equation 13- General Error Correction Model.

## Equation 14- Error Correction Model for Country Specific equation 7.

## Equation 15- Error Correction Model for Country Specific equation 8.

## Equation 16- Error Correction Model for Country Specific equation 9.

## Equation 17- Error Correction Model for External Factor equation 10.

## Equation 18- Error Correction Model for External Factor equation 11.

## Equation 19- Error Correction Model for External Factor equation 12.

## Equation 20- Engle Granger Cointegration

## Equation 21- Long Run Equilibrium.

## Equation 22- Unit Root Equation.

## Equation 23- Cointegrating Regression Durbin Watson.

## Equation 24- Trace Statistic Cointegration.

## Equation 25- Maximum Eigenvalue Cointegration.

## Equation 26- Causality Equation for Variable Y.

## Equation 27- Causality Equation for Variable X.

## Equation 28- Wald F-Test Framework.

## Equation 29- pth order of Autoregressive Conditional Heteroscedasticity.

## Equation 30- Conditional Mean of GARCH model.

## Equation 31- Conditional Variance of GARCH model.

## Equation 32- Return of Oil Price.

## Equation 33- Return of Oil Price.

## Equation 34- Logarithm of Oil Price.

## Equation 35- Logarithm of Return of Oil Price.

## Equation 36- Conditional Mean of EGARCH model

## Equation 37- Conditional Variance of EGARCH model.

## Equation 38- Test of Significance.

## Equation 39- F-Statistics test.

## Equation 40- Coefficient of Determination (R2).

## APPENDIX D: Robustness Checks

Serial Autocorrelation: The Breusch Godfrey tests for serial correlation is employed.

## Remedy: Newey West Standard Errors.

Multicolinearity: The variables in the equations are all endogenous variables based on the vector autoregressive system.

## Remedy: Tests for Cointegration. Vector Error Correction Model.

Heteroscedasticity: This is not a common feature with time series data, however the Newey West Standard Errors are an extension of the White Heteroscedascity Corrected Standard errors.

## Remedy: Newey West Heteroscedasticity Autocorrelation Consistent (HAC) Standard Errors.

## Bollersev and Woodridge Robust Standard Errors.

## APPENDIX E: Critical Values

## Critical Values for Augmented Dickey Fuller Tests.

## Significance Level

## 10%

## 5%

## 1%

CV for constant but no trend

-2.57

-2.86

-3.43

CV for constant and trend

-3.12

-3.41

-3.96

## Critical Values for Kwiatkowski-Phillips-Schmidt-Shin Tests.

Distribution

Upper tail percentiles

## 1

## 5

## 2.5

## 1

0.3

0.4

0.5

0.7

0.1

0.1

0.1

0.2

## Critical Values Durbin Watson Cointegration Tests.

## Significance Level

## 10%

## 5%

## 1%

## D

## 0.322

## 0.386

## 0.511

## APPENDIX F- Abbreviations in this Study

## FSI- Financial Soundness Indicators.

## IMF- International Monetary Fund.

## CBN- Central Bank of Nigeria.

## NBS- National Bureau of Statistics.

## EMDB- Emerging Markets Database.

## OPEC- Organization of Petroleum Exporting Countries.

## OECD- Organization for Economic Co-operation and Development.

## GCC- Gulf Cooperation Council

## ADF- Augmented Dickey Fuller.

## PP- Philips Perron.

## KPSS-Kwiatkowski-Phillips-Schmidt-Shin.

## CRDW- Cointegrating Durbin Watson.

## OLS- Ordinary Least Squares.

## ARIMA- Autoregressive Integrated Moving Average.

## LLF-Log Likelihood Function.

## EG- Engle Granger

## JC- Johansen Cointegration.

## VECM- Vector Error Correction Model.

## VAR-Vector Autoreggressive

## ARCH- Autoregressive Conditional Heteroscedasticity.

## GARCH – Generalized Autoregressive Conditional Heteroscedasticity.

## EGARCH- Exponential Generalized Autoregressive Conditional Heteroscedasticity.

## BIBLOGRAPHY

## Adamu, A. (2007), ‘The Effects of global financial crisis on Nigerian Economy’, International Journal of Investment and Finance, Vol. 1, pp. 1-2.

Adegbite, O. Ayadi, F. and Ayadi, S. (2008), ‘The impact of Nigeria's external debt on economic development’ , International Journal of Emerging Markets, Vol. 3(3), pp. 285 – 301.

Aliyu, S. (2009), ‘Impact of Oil Price Shock and Exchange Rate Volatility on Economic Growth in Nigeria, An Empirical Investigation’, Research Journal of International Studies , Vol. 1(11), pp. 4-15.

Askari, H. and Krichene, N. (2006), ‘Oil price dynamics’, Energy Economics,

Vol. 30(5), pp. 2134-2153.

Baker, M. and Wurgler, F. (2002), ‘Market Timing and Capital Structure’, Journal of Finance, Vol. 22(1), pp. 2-32.

Balke, N. Brown, S. and Yücel, K. (2002), ‘Oil Price Shocks and the U.S. Economy:

Where Does the Asymmetry Originate?’ , The Energy Journal, Vol. 23(3), pp. 27-52.

Bollerslev, T., (1986), ‘Generalised autoregressive conditional Heteroscedasticity’, Journal of Econometrics, Vol. 31, pp. 307–327.

Brooks, C. (2008), Introductory Econometrics for Finance, Cambridge: Cambridge University Press.

Burbridge, J. and Harrison, A. (1984), ‘Testing for the Effects of Oil-Price Rises Using Vector Autoregressions’, International Economic Review, Vol. 25(2), pp. 459-84.

Chen, R. Roll, S. Ross, S. and Nai-Fu. V. (1986), ‘Economy and Stock Markets’ , Journal of Business, Vol. 59(3) pp 189-236.

Cologni, A. and Manera, M. (2007), ‘The asymmetric effects of oil shocks on output growth: A Markov–Switching analysis for the G-7 countries’, Economic Modelling, Vol. 26(1), pp. 1-29.

Cologni, A. and Manera, M. (2007), ‘Oil prices, inflation and interest rates in a structural cointegrated VAR model for the G-7 countries’, Energy Economics, Vol. 30(3), pp. 856-888.

Cunado, J. and Perez, F. (2005), ‘Oil prices, economic activity and inflation: evidence for some Asian countries’, The Quarterly Review of Economics and Finance, Vol. 45(1), pp. 65-68.

Dickey, D. and Fuller, W. (1979) ,’ Distribution of the Estimators for Autoregressive Time Series with a Unit Root’ , Journal of the American Statistical Society, Vol. 75, pp. 427-431.

Faff, R. and Nandha, A. (2007), ‘Does oil move equity prices? A global view’, Energy Economics, Vol. 30(3), pp. 986-997.

Faff, R. and Brailsford, T. (1999), ‘Oil price risk and the Australian stock market’, Journal of Energy Finance and Development, .Vol. 4(1), pp. 69-87.

## Fama, F. and French, R. (2002), ‘Testing Trade-Off and Pecking Order Predictions About Dividends and Debt’, The Society for Financial Studies, Vol. 4(15), pp. 1-33.

Farooq, M. and Bradley, T. (2009), ‘Volatility transmission between oil prices and equity sector returns’ , International Review of Financial Analysis, Vol. 18(3), pp. 95-100.

Ferderer, P. (1996), ‘Oil Price Volatility and the Macoreconomy’, Journal of Macroeconomics, Vol.18(1), pp. 1-26.

Frank, M. and Goyal, K. (2003), ‘Trade-Off and Pecking Order Theories of Debt’, Handbook of Empirical Corporate Finance, pp. 135-202.

George, F. (2010), ‘Macro economy, stock market and oil prices: Do meaningful relationships exist among their cyclical fluctuations?’, Energy Economics,

Vol. 32(4), pp. 877-886.

Gisser, M. and Goodwin, H. (1986), ‘Crude oil and the macoreconomy: Tests of some popular notions’ , Journal of Money Credit Banking, Vol.18 (1), pp. 95-103.

Granger, J. (1969), ‘Investigating causal relations by econometric models and cross-spectral methods’, Econometrica, Vol. 37 (3), pp. 424–438.

Gujarati, D. and Porter, D. (2009), Basic Econometrics, Singapore: McGraw Hill Education.

Engle, F. (1982), ‘Autoregressive conditional Heteroscedasticity with estimates of the variance of the UK inflation’ , Econometrica, Vol.50, pp .987-1008.

Hamilton, D. (1983), ‘Oil and the Macoreconomy since World War II’, Journal of

Political Economy, Vol.91, pp. 228-248.

Hamilton, D. (1996), ‘This is what happened to the oil price-macoreconomy relationship’ ,

Journal of Monetary Economics, Vol.38, pp. 215-220.

Hamilton, D. (2000), ‘What is an oil shock?’, NBER Working Paper 7755.

Hammoudeh, S. and Choi, K. (2005), ‘Characteristics of permanent and transitory returns in oil-sensitive emerging stock markets: The case of GCC countries’, Journal of International Financial Markets, Institutions and Money, Vol.17(3), pp. 231-245.

Harris M. and Raviv, A. (1990), ‘Capital Structure and the Informational Role of Debt’, The Journal of Finance, Vol. 45(2), pp. 321-349.

Harvey, C. (1995), ‘Predictable Risk and Returns in Emerging Markets’, The Review of Financial Studies, Vol. 8(3), pp. 773-816.

Hooker, M. (1996), ‘What happened to the oil price-macoreconomy relationship?’ ,

Journal of Monetary Economics, Vol. 5(38), pp. 195-213.

Hovakimian, A. (2004), ‘The role of target leverage in security issues and repurchases’. Journal of Business, Vol. 77(4), pp 25-35

Irwin, F. and Lang, H. (1988), ‘The size effect on stock returns: Is it simply a risk effect not adequately reflected by the usual measures?’, Journal of Banking and Finance, Vol. 12(1), pp. 13-30.

Ito, K. (2008), ‘Oil price and macoreconomy in Russia’, Economics Bulletin, Vol. 17(17), pp. 1-9.

John, T. (1988), ‘Are the macroeconomic effects of oil-price changes symmetric?’,

Carnegie-Rochester Conference Series on Public Policy, Vol 28, pp. 325-368.

Jones, C. and Kaul, G. (1996), ‘Oil and the Stock Markets’, The Journal of Finance, Vol LI( 2), pp. 115-136

Kennedy, P. (1999), A Guide to Econometrics, Oxford: Blackwell Publishers Inc.

Kwiatkowski, D. Phillips, B. Schmidt, P. and Shin, Y. (1992), ‘Testing the Null Hypothesis of Stationary against the Alternative of a Unit Root’, Journal of Econometrics, Vol. 54, pp. 159-178.

Lee, K. Ni, S. Ratti, R. (1995), ‘Oil Shocks and the Macoreconomy: The Role of Price Variability’, Energy Journal, Vol. 16, pp. 39-56.

Lee, K. and Ni, S. (2002), ‘On the dynamic effects of oil price shocks: A study using industry

level data’, Journal of Monetary Economics, Vol. 49, pp. 823–852.

Marsh, P. (1982), ‘The Choice Between Equity and Debt: An Empirical Study’, The Journal of Finance, Vol. 37(1), pp. 121-144.

Miguel, C. Manzano, B. and Martin-Moreno, H. (2003), ‘Oil Shocks and the Business Cycle in Europe’ , Universidad de Vigo and rede, pp. 2-16.

## Mishkin, F. (2009), ‘Is Monetary Policy Effective During Financial Crises?’, American Economic Review, American Economic Association, vol. 99(2), pp. 573-77.

## Mork, K. Olsen, O. and Mysen, T. (1994), ‘Macroeconomic responses to oil price increases and decreases in seven OECD countries’, Energy Journal, Vol. 15(4), pp. 19-35.

Mork, K. (1989), ‘Oil and the Macoreconomy, When prices Go Up and Down: An Extension of Hamilton’s Results’ , Journal of Political Economy, Vol. 97(51) pp. 230-267

Moshirian, F. (2009), ‘The current global financial crisis and the evolution of the international financial institutions’ . A Paper to be Presented at the 22nd Australasian Finance and Banking Conference, Sydney, Australia.

## Mussa, M. (2000), ‘The impact of higher oil prices on the global economy’, The International Monetary Fund.

## Myers, C. and Lakshmi, S. (1994), ‘Testing Static Trade-off Against Pecking Order Models of Capital Structure’, Journal of Financial Economics, Vol. 51(2), pp. 219-244.

Narayan, P. and Narayan, S. (2007), ‘Modelling oil price volatility’, Energy Policy, Vol. 35(12) , pp. 6549-6553.

Nelson, D. (1991), ‘Conditional Heteroscedasticity in asset return: a new approach’. Econometrica, Vol. 59, pp. 347–370.

Nina, B. and Sweder, W.(2008), ‘Managing Oil Revenue Volatility in Nigeria: The Role of Fiscal Policy, World Bank Report, Chapter 10 pp. 427-460.

Nzotta, M. and Okereke, J. (2009), ‘Financial Deepening and Economic Development of

Nigeria: An Empirical Investigation’, Journal of Accounting, Economics, Finance and Banking Research, Vol. 5(5), pp.

Oliver, E. and Williamson, E. (1988), ‘Corporate Finance and Corporate Governance’, The Journal of Finance, Vol. 43(3), pp. 567-591.

Olomola, A. and Adejumo V. (2006) ‘Oil Price Shock and Macroeconomic Activities in

Nigeria’, International Research Journal of Finance and Economics, (3). pp. 28- 34

Omisakin, O. Adeniyi, O. and Omojolaibi, A. (2009), ‘A Vector Error Correction Modelling of Energy Price Volatility of an Oil Dependent Economy: The Case of Nigeria’, Pakistan Journal of Social Sciences, Vol 6. (4) , pp. 207-213.

Oskooe, O. (2010), ‘Emerging Stock Market Performance and Economic Growth’, American Journal of Applied Sciences, Vol.7 (2), pp. 265-269.

Papapetrou, E. (2001), ‘Oil price shocks, stock market, economic activity and employment in Greece’, Energy Economics, Vol. 23 (5), pp. 511-532.

Phillips, P. and Perron, P. (1988), ‘Testing for a unit root in time series regression’ Biometrika Vol.75, pp. 335-346.

Rafiq, H. Salim, R. and Bloch, H. (2008), ‘Impact of crude oil price volatility on economic activities: An empirical investigation in the Thai economy’, Resource Policy, Vol 34 (3), pp. 121-131.

Rajan, R. and Zingales, L. (1995), ‘Is there an Optimal Capital Structure? Some Evidence from International Data’, Journal of Finance, Vol. 50, pp. 1421-60.

Ramanathan, R. (2002), Introductory Econometrics with Application, Ohio: Thomson Learning.

Rotemberg, J. Woodford, M. (1996), ‘Imperfect Competition and the Effects of Energy Price Increases on Economic Activity’, Journal of Money, Credit, and Banking, Vol. 28(4), pp. 550-577.

Sadorsky, P. (1999), ‘Oil price shocks and stock market activity’, Energy Economics , Vol. 21, pp. 449-469.

Sadorsky, P. (2003), ‘The macroeconomic determinants of technology stock price volatility’, Review of Financial Economics, Vol. 12, pp. 191-205.

Sadorsky, P. (2006), ‘Modelling and forecasting petroleum futures volatility’, Energy Economics, Vol. 28, pp. 467-488.

Sari, U. and Soytas, R. (2006), ‘Energy consumption and income in G-7 countries’, Journal of Policy Modelling, Vol.28(7), pp. 739-775.

Schubert, S. Turnovsky, S. (2009), ‘The impact of oil prices on an oil-importing developing economy’. Journal of Development Economics, Vol. 5(1), pp. 122-130.

Seddighi, K. Lawler, A. and Katos, V. (2000), Econometrics; A Practical Approach, London: Routledge Publishing.

Smith, C. and Watts, R. (1992), ‘The Investment Opportunity Set and Corporate Financing, Dividend, and Compensation Policies’, Journal of Financial Economics, Vol. 32 pp. 263-292.

Studenmund, A. (2006), Using Econometrics; A Practical Guide, Adelaide: Pearson Education Inc.

Titman, S. and Wessels, R. (1988), ‘The Determinants of Capital Structure Choice’, The Journal of Finance, Vol. 43(1), pp. 1-19.

Wiwattanakantang (1999), ‘Controlling Shareholders and Corporate Value: Evidence from Thailand’, Pacific Basin Finance Journal, Vol.1, pp. 5-10.

Yang, C. Hwang, M. and Huang, B. (2002), ‘An analysis of factors affecting price volatility of the US oil market’, Energy Economics, Vol. 24, pp. 107-119.

## Need help with your literature review?

Our qualified researchers are here to help. Click on the button below to find out more:

## Related Content

In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: