This action might not be possible to undo. Are you sure you want to continue?

SAMUEL DAGADU

Subject Area: Finance & Economics

Submitted: February 2010

Dissertation submitted to University of Leicester in partial fulfilment of the requirements for the degree of Master of Science in Finance.

Table of Contents

Acknowledgments Executive Summary 1.0 Introduction 1.1 1.2 1.3 Macroeconomic factors and Stock market returns The Role of Governments The Economy of Ghana 1.3.1 1.3.2 1.3.3 1.3.4 1.3.5 1.4 1.5 1.6 1.7 1.8 2.0 2.1 The Recent Developments Oil Revenue Expectation Macroeconomic Trends (2000 to 2009) Economic Policies Policy Implementation Experience 9 9 10 11 11 12 13 17 18 19 21 22 23 24 25 25

The Ghana Stock Exchange Other Institutions of Importance: The Bank of Ghana Research Questions Interest in Research Structure of Dissertation Macroeconomic factors that affect Stock Market Returns 2.2 Models used to establish relationship between

Literature Review

macroeconomic factors and stock returns 2.3 2.4 Previous research on Ghana Stock Exchange Theoretical considerations 2.4.1 Interest rates 2.4.2 Inflation 2.4.3 Fiscal deficits/surpluses 2.4.4 Exchange rates 2.4.5 Gross Domestic Product (GDP)

27 29 30 30 31 31 32 33

2

2.4.6 The Role of Government

35

3.0

Methodology 3.1 3.2 3.3 3.4 Data Collection Variable selection Model specification Tests conducted 3.4.1 Visual Inspection 3.4.2 Unit Roots Tests 3.4.3 Co integration Tests 3.4.4 Vector Error Correction 3.4.5 Granger Causality Tests 3.5 3.6 Data description Regression

37 37 37 39 40 40 40 41 42 42 45 45 45 46 47 48 48 48 49 49 50 53 54 55 58 61 62 3

3.7.0 Hypothesis 3.7.1 Interest rates 3.7.2 Inflation 3.7.3 Exchange rate 3.7.4 Fiscal deficits/surpluses 3.7.5 Gross Domestic Product 4.0 Data Analysis 4.1 4.2 4.3 4.4 5.0 6.0 7.0 8.0 Unit Roots Test Co-integration Vector Error Correction Granger Causality

Research Findings Conclusion Recommendation Reflection

9.0

References

65 70-103

10.0 Appendices

ACKNOWLEDGEMENTS

I would like to thank Dr. Tomasz Wisniewski of University of Leicester, for his guidance and advice on topic selection. I also thank Dr. Tse, University of Leicester, for his advice, guidance and support from the proposal stage through to final submission. My appreciation also goes to my employer, Social Security and National Insurance Trust for sponsoring my studies, to Mr. Kwasi Boaten - Former Director General and Sheila Sampson, Training Manager, for their interest in my studies. My appreciation and thanks goes to my Superiors, Colleagues and Subordinates for their support during the period of my studies. I thank Dr. E. E. Y Dagadu for his support and encouragement throughout my career development and for buying the econometrics software used for the analysis. I finally wish to thank my immediate family Bridget, Teddy, Jeffrey, Jennifer, Harold and Philip for their support, patience and understanding during the period of my studies. It would have been impossible to complete my dissertation without their involvement. I owe a debt of gratitude to Mr. Simon Dewotor and his family for mentoring me and showing me the path to success.

4

Executive Summary This study examines the long and short run relationships between

macroeconomic variables and stock market returns in general using Ghana Stock Exchange Index as a special case, to find answers to the following questions: What macroeconomic factors drive the performance of Ghana Stock Index? How does GDP, Fiscal balance, Inflation, Interest rates and Exchange rates impact the Ghana Stock Index (GSI)? The macroeconomic variables used for this study are Gross Domestic Product (GDP), Fiscal balance (deficit/surpluses), Inflation, Interest rates and Exchange rates. Finding answers to these questions will help add to existing knowledge about the underlying causes of price movements of the Ghana Stock Exchange and how these variables can be useful in predicting the performance of the Ghana Stock Exchange. This study was undertaken in partial fulfilment of the requirements for the award of degree of Master of Science in Finance, University of Leicester. There is no economic theory that explains the linkage between macroeconomic variables and stock market performance in one direction, but there are several macroeconomic factors that have been identified as having impact on stock market performance. This study used the Dividend Discount Model to select the variables considered in the study.

5

Cointegration, Error Correction and Granger causality techniques were used to establish relationships. The long term relationships were analysed and established by Johansen and Juselius Multivariate Co integration Approach. Short term relationships were analysed and established through Vector Error Correction models and Granger Causality tests. Monthly time series data from January 1991 to December 2008 were used. Data on Fiscal balance, Inflation, 91 day Treasury bill rates and Exchange rates movements were obtained from the Bank of Ghana and stock Index movements were obtained from the Ghana Stock Exchange. Empirical findings revealed that macroeconomic variables considered are cointegrated and have statistically significant coefficients that indicate the existence of long term relationships. The strength of relationship in the short run is however weak indicating the variables do not have short term relationship. The results indicate that the Ghana Stock Market has significant positive long run relationship with GDP and Inflation but negative long run relationship with fiscal deficit, interest rates and exchange rates. These results are consistent with current theoretical arguments regarding these relationships. The results are in line with anticipated relationships between the variables, with the exception of Inflation, which shows a positive instead of negative relationship. The policy relevance of this study is significant now and into the future because the economic policy of Ghana is geared towards growth and stability. The Country is expecting the production and export of crude oil. GDP is projected to grow from 5% in 2010 to 24.2 % in 20111. The commencement of oil production will create serious macroeconomic imbalance which needs to be managed carefully to ensure sustainable growth and stability. Moreover, individuals, Pension Funds and Institutional Investors invest on the Stock market to ensure a future stream of income. Any erosion of capital in the market due to

1

IMF Country Report No.9/256, August 2009 6

macroeconomic imbalances can have dire consequences for all market participants. This study will be useful in formulating both fiscal and monetary policies. Other studies on the Ghana Stock Exchange reported on the impact of inflation, interest rates, foreign direct investments and exchange rates. No study considered the effect of GDP and Fiscal imbalances on the Ghanaian Stock Index. This paper is therefore adding to the body of knowledge on the relationship between these macroeconomic variables and stock market returns in Ghana. The lag length used for the error correction in this study may be a limitation and might have accounted for the disparity in results when this study is compared with other studies conducted on the GSI. The slow pace of adjustment suggests that the macroeconomic factors examined in this study are not exhaustive. There may be other significant factors which were not considered in this study. Another limitation is the lack of monthly data for some of the variables. This may be a contributory factor in the lag length required for the variables to relate to each. The Government of Ghana is concerned about stability and growth. Greater stability can be achieved if the Government of Ghana establishes the following through the Bank of Ghana and the Ministry of Finance. Stable monetary growth to match the future growth potential of the economy; Manage money supply with the hope of maintaining prices; Maintain fiscal deficit or surpluses during periods of recession and periods of expansion respectively; Automatic stabilizers like unemployment benefits during periods of low economic activities, high corporate taxes during booms and low corporate taxes during recession, and wider coverage and progressive personal income tax regimes. The use of Oil revenue should also be regulated to avoid negative impact on macroeconomic variables.

7

This study has contributed to existing knowledge on the grounds that there is a lag effect of changes in macroeconomic factors on the Ghanaian Economy in general and the Stock Exchange in particular. The transmission mechanism of policy needs further study to identify reasons for the lag and what can be done to correct it. For example, the current debate on radio and in the print media is for the Government to compel Banks to reduce their lending rates in line with reduction in inflation, prime rates and general improvement in other macroeconomic factors.

8

1.0

Introduction

The Ghana Stock Exchange experienced a highly volatile performance since it was established in 1990. It was pronounced the best performing stock exchange in the World in 2004. It performed above 100% in the following years 1993, 1994, 2003 and performed below 0% (zero) in the following years 1990, 1991, 1992, 1999, 2005 and 2009. The causes of these fluctuations in performance can be attributed to the changes in macroeconomic factors from theoretical and empirical point of view but not by merely looking at the movement of macroeconomic variables. The GSE performance is measured by GSE AllShare- Index, which is a market value weighted index. 1.1 Macroeconomic factors and Stock market returns

Macroeconomic variables have significant effect on stock price movement and returns but the basis of the causal relationship between macroeconomic variables and stock prices is not known with certainty as indicated by Flannery and Protopapadakis (2002). Efficient market hypothesis attributes movement of stock prices to new information that affect the expected discount rates or future income. An efficient market, for example, incorporates all current market information in stock prices. Any new information is captured instantaneously to reflect all available information. Studies by Fama and Schwert (1977), and Jaffe and Mandelker (1976) suggest that new information on macroeconomic factors have an impact on stock prices. This affirms the belief that macroeconomic variables influence stock returns and thus proposes that stock markets are not

9

efficient. Empirical evidence suggests that stock returns respond to monetary news but there is no theory on the relationship of stock returns with macroeconomic variables in one direction, even though stock prices are known to react to market forces. Uncertainty remained about relationships between macroeconomic variables and stock performance because of varying economic conditions of nations, different data set and different testing methods used to establish these relationships. Economic factors that impact on changing investment opportunities; the pricing polices; and factors which affect dividends theoretically, impact pricing and performance of stock exchanges. Predictability of stock market returns using macroeconomic factors suggests that markets are not efficient. As indicated by Fama (1991), stock prices reflect expectations of earnings, dividends, interest rates and future economic activity. If stock prices reflect the underlying fundamentals, then we can say that there is a causal relationship between macroeconomic variables and stock prices. Participants in the stock market anticipate real returns from the stock market so stock prices will move directly with inflation. Investors directly compare earning yield on stocks with treasury yields and move funds from one market to the other, an inverse relationship between stock returns and interest rates is expected. Fiscal deficits increase Government borrowing which increases interest rates, which crowd out private individuals and businesses, denying listed companies of much needed capital resulting in low returns. 1.2 Role of Governments

Twerefou and Nimo (2005) reported that businesses factored into their operations Government macroeconomic targets for the year. “This invariably feeds into the determination of stock prices of listed companies”. Twerefou and Nimo (2005:169). Government’s fiscal and monetary policies have significant effect on the economy and therefore the capital market. Fiscal policy is aimed at consumer demand in an effort to manage economic growth. Tax cuts encourage

10

demand and for that matter spending. Tax increases discourage spending and slow down the economy. Government’s debt financing causes interest rates to rise and this directly causes Inflation and depreciation of the local currency. Depreciation results in capital flights. Investors who do not have confidence in the economy divert resources from the long term investments to short term treasury bills and consumables or real estate. Listed Companies are starved of needed resources to finance viable projects, since they cannot borrow at the prevailing high interest rates, they become less competitive and their profit levels fall leading to a fall in returns. The Government of Ghana has a key role to play in ensuring a sound capital market through macroeconomic measures such as the fiscal environmenttaxation; legal, regulatory and institutional infrastructure, as well as monetary policies. Other factors like oil price hikes, change of Governments, international financial and economic development/crises also have impact on the economy but these factors were not considered in this study. 1.3 The Economy of Ghana

1.3.1 The Recent Developments Ghana went through a keenly contested democratic election in December 2008. The National Democratic Party took over from the New Patriotic Party, which was in Government for the past eight (8) years, in January 2009. The new Government adopted a program of macroeconomic stabilization and growth through reforms in joint partnership and support with the International Monetary Fund (IMF).

11

The real GDP growth somewhat consistently increased from 3.7% in 2000 to 7.3% in 2008 on the back drop of significant debt relief and strong commodity prices which gave some fiscal relief to the country in its developmental efforts, leading to higher output, declining Inflation, and poverty reduction. The improved macroeconomic environment enabled Ghana to raise US$750 million in Eurobonds in October 2007 for infrastructural development in the energy sector in particular. Important institutional reforms have been undertaken in the financial sector to safeguard the stability of the financial system but the nation suffered a severe energy crisis in 2006-2007 as a result of prolonged drought, which lead to near shutdown of Akosombo Hydro Electric generating plant. This resulted in a shift in over dependence on hydro power to thermal power at the time of rising crude oil prices creating macroeconomic pressure on the economy. The global food and fuel price increases of 2007-2008 hit Ghana very hard. The government implemented some social mitigation policies which dampened the effects of the global food crises. 1.3.2 Oil Revenue Expectation Ghana discovered crude oil along Cape Three Points in fields named the Jubilee fields. The dominant players are Tullow Oil Plc and Kosmos Energy Plc Companies. The Ghana National Petroleum Company (GNPC) is also in collaboration with some Chinese and other oil exploration companies that are prospecting for oil and gas in five sedimentary basins including inland Voltaian, offshore Tano and Saltpond basins. It is expected that drilling in commercial quantities from the Jubilee fields will commence by the last quarter of 2010. Daily production is expected to be 120,000 barrels until the end of 2012 when the expected production target is 250,000 barrels per day. The Government of Ghana is making efforts to create policies and frameworks that will ensure maximum gains from the oil industry. Particular concerns are how much to save,

12

how much to spend on what, in order to insulate the economy against fiscal fluctuations, and to protect the economy against exchange rate appreciation. Other concerns are environmental protection, technological transfers and the avoidance of the “Dutch disease”. The projected direct annual oil revenue is 7% of 2008 GDP and about 26% of 2008 domestic revenue. IMF projects 4% to 5% of annual revenue as indicated by ISSER Report (2008). This is indeed a substantial injection. It however, excludes revenue from Gas, which is currently controlled domestically, and downstream industries that will spring up as a result of the oil Industry. 1.3.3 Macroeconomic Trends (2000-2009)

Ghana experienced macroeconomic instability for many years in the past. This affected growth of the economy. The Country achieved significant gains in the macroeconomic and social sectors from 2000 to 2005. The past four (4) years however experienced imbalances, caused by the energy crises of 2006-2007 and external shocks resulting from rising food and oil prices. This was compounded by the 2008 general elections because of increased direct and indirect election related expenses and energy related subsidies, higher wages and salaries among others. All these culminated in raising fiscal deficits from 7.81% of GDP in 2006 to 11.48% in 2008 and 10.7% in 2009. The global financial crises contributed to balance of payment pressures in the face of reducing private remittances and outflow of portfolio investment proceeds. The exchange rate plummeted from mid 2008 to June 2009. The rate of depreciation slowed down because of injection of foreign currency into the economy by IMF. The cedi weakened against all the major trading currencies like the Dollar, the Pound Sterling and the Euro. The Bank of Ghana set a minimum capital requirement of GH¢60 million for Banks by the end of 2009. Domestic Banks are however required to increase

13

their capital to GH¢25 million by the end of 2010 and then to GH¢60 million by the end of 2012. The National Pensions Act of 2008 came into effect in January 2010; it provided the frame work for a three tier pension system. The first tier is run by Social Security and National Insurance Trust (SSNIT). The Occupational Pension Schemes and Private Pension schemes are run by Trustees who appoint Fund Managers and Custodians. They all operate under the regulatory regime of a Pensions Authority. The Pension Authority is expected to ensure that Pension funds are prudently managed. It is expected that there will be an injection of capital in the system for speculative and investment activities. Table 1 shows the trend in some macroeconomic variables from 1993 to 2009. Table 1 Trends of Macroeconomic variables & Chart 1

Market Returns% Real GDP% Fiscal bal as % of GDP

91 day T/Bills

YEAR 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Interest%

Inflation%

%Δ GHS/ USD

113.74 124.34 6.33 13.82 41.85 69.19 (15.22) 16.55 11.42 45.96 154.67 91.33 (29.85) 4.97 31.84 58.06 (46.52) 4.7*

5.00 3.30 4.00 4.60 4.20 4.70 4.40 3.70 4.20 4.50 5.20 5.60 5.90 6.40 5.70 7.30

10.7*

(2.65) 2.26 0.95 (3.16) (8.21) (6.07) (6.51) (8.62) (4.36) (6.11) (3.52) (2.77) (1.95) (7.81) (8.10) (11.48)

32.00 28.38 36.50 41.98 42.73 34.28 26.36 39.10 40.99 25.10 28.80 17.28 15.43 10.16 9.92 17.79 23.52

70.80 32.70 20.80 15.70 13.80 18.19 4.94 40.20 43.49 9.49 29.76 18.18 15.48 10.96 10.72 16.46 19.30

(13.13) 155.11 (27.46) 36.28 25.21 12.89 15.21 104.61 31.43 10.62 9.65 3.53 0.77 1.12 1.98 13.05

18.11

Source: Bank of Ghana & Ghana Stock Exchange * Estimates.

14

160

120

80

40

0

-40 1992

1994

1996

1998

2000

2002

2004

2006

2008

M/Returns% Fiscal% Inflation%

GDP% Interest% Dollar%

15

M/Returns%

160 120 80 40 0 -40 92 94 96 98 00 02 04 06 08 8 7 6 5 4 3 92 94 96 98

GDP%

00

02

04

06

08

Fiscal%

4 50 40 30 -4 20 -8 10 0 92 94 96 98 00 02 04 06 08 92 94 96

Interest%

0

-12

98

00

02

04

06

08

Inflation%

80 160 120 80 40 40 20 0 -40 92 94 96 98 00 02 04 06 08 92 94 96 98

Dollar%

60

0

00

02

04

06

08

16

1.3.4 Economic Policy The Government of Ghana is pursuing economic policies that will ensure the attainment of macroeconomic stability and growth. The strategy is being accomplished through fiscal discipline and programs geared towards prudent public expenditure management, enhanced domestic revenue mobilization, enforcement of public procurement laws and restructuring of utility companies to reduce subsidy on the consolidated budget from these utilities. Public sector reforms are also being under taken to bring on board a single spine salary structure. The Government intends to reduce the budget deficit through cuts in low priority public expenditures in order to reduce total public expenditure in relation to GDP. Revenue mobilization was to be strengthened to increase revenues in relation to GDP. Monetary Policy support Government’s fiscal consolidation efforts with emphasis on price stabilization and growth as well as exchange rate expectation. This is done through inflation targeting which is a framework by which policies are guided by the expected path of future inflation relative to already planned and announced inflation target. The Ghana Poverty Reduction Strategy (GPRS II) has at its base the provision of a “life line” scheme to mitigate the risks non income earners or low-income vulnerable groups face. The Livelihood Empowerment Against Poverty (LEAP) cash transfer program, school feeding and free maternal care programs, capitation grant and youth in employment programs, provision of free exercise books and free school uniforms for school pupils in deprived communities are examples of mitigating factors in place.

17

1.3.5 Policy Implementation Experience Maintaining fiscal discipline has been and is still a major challenge for all Governments of Ghana. Ghana has a well thought out fiscal and monetary policy but the policy transmission mechanism has some challenges largely because of implementation. Implementation of economic policy is difficult because of lags between the time policy makers identify the problem and when decisions are made as well as when the decisions are implemented as against when the full impact is felt, the impact of unanticipated changes is greater than that of anticipated changes. Inflation targeting for instance is based on expected future inflation. This may suffer from recognition lag, which is the time it takes for policy makers to recognize that an economic change which needs a policy change has occurred. Inflation for example can assume a downward trend while discount rates and particularly Bank base rates will either remain stable or be on the increase as the economy of Ghana experienced in the past and currently. Implementation time lag is shorter for monetary policy than fiscal policy. The Monetary Policy Committee of Bank of Ghana meets frequently to determine the Prime rate on which interest rates and inflation targeting is based, but fiscal policy needs approval from Parliament. The impact lag is however shorter for fiscal policy than monetary policy. Economic policy based on past events creates certain errors depending on whether the policy is based on adaptive expectation or on rational expectation. Adaptive expectation yield systematic errors resulting in over or under estimation. Inflation targeting policy is based on rational expectation, considers the future and currently available information instead of historical data. Random forecasting errors are made but the problem is the time lags and whether the policy change is expected or not. Expected monetary and fiscal policies have little or no impact

18

on the economy as these factors might have been factored in prices or budgets of companies. Unanticipated monetary policy changes have a greater impact. 1.4 The Ghana Stock Exchange

The Ghana Stock Exchange was established in July 1989 as a company limited by guarantee under the Companies code 1963. In October 1990, the Exchange was given recognition as an authorized Stock Exchange under the Stock Exchange Act of 1971 (Act 384). Trading commenced on the floor on 12th November 1990. The status of the Company changed from private to public company limited by guarantee in April 1994. The GSE is governed by a Council of representatives from licensed dealers, listed companies, banks, insurance companies, the money market and the general public. The exchange has been trading daily since mid 2008. . Prior to that, there were three (3) dealing dates every week. Trading activities are no longer done on the trading floor. A central securities depository has been established, securities have been dematerialized and trades are now done and settled electronically from Brokers offices. Dealers and the Investing public are required to register their shares online to be able to trade. The listing requirements include capital adequacy, profitability, efficiency of management, and float of shares and years of operational existence. The GSE performance is measured by GSE All- Share- Index, which is a market value weighted index. The GSE currently has 35 listed companies, one depository share and one preference share trading on the exchange actively. Two corporate Bonds, HFC and Standard Chartered Bank (SCB) Medium Term Bonds are also listed on the exchange. The HFC bonds will mature in March 2012. These bonds are dollar denominated with coupons priced at 6 months USD Libor +100bp. SCB Bonds matured in December 2009. The Government of Ghana has three categories of bonds, trading with terms of two (2) years, three (3) and five (5) years. The coupon rates range from (12.8%-21%), (12.08%-16%), and (13.67%-15%)

19

respectively. The current yield on the other hand ranges from (12.64% to 17.11%) for 2 year bonds and (14.87% to 17.44%) for three year bonds. The listed shares are categorized into Manufacturing, Financial, Mining and Gas & Oil sub-sectors. The Bank of Ghana and Securities Exchange Commission are the regulators. The Ghana Stock Exchange achieved recognition in the global investment arena. In 2003 it achieved performance of 154.67% and was recognized as the best performing market in 2004. The remarkable performance was attributed to economic performance resulting from stable and good macroeconomic factors during the period leading to investor interest on the exchange. In 2008, Ghana Stock Exchange was adjudged one of the best during the period of financial meltdown of advanced markets. The remarkable performance was attributable to economic performance resulting from stable and strong macroeconomic factors. The feat of 2008 was to be followed by over 46% negative performance in 2009, the lowest in Africa. This poor performance was also attributed to poor macroeconomic factors by Market Analyst, a critical look at the performance figures suggests some lag of the effect of Macroeconomic factors on market performance. The good performance of the economy in 2005 to 2006/7 impacted the market in 2008. The poor performances in 2007 and 2008 impacted the market in 2009. It is difficult to say whether the Ghana Stock Exchange is efficient or not. It is difficult to depend on past price movement to make gains but the same cannot be said about the release of information. Some Companies announce good returns and prospects but nothing happens to their prices. This study does not consider the efficiency of the exchange.

20

1.5

Institutions of Importance :The Bank of Ghana

The Bank of Ghana (BOG) is an independent body with oversight responsibility over the monetary policy and strategy of Ghana. The objectives of BOG include: Maintenance of the general price levels through implementation of monetary policy. Support the general economic policy of Government and promote economic growth Regulation of the financial sector of the economy The goals of monetary policy of Ghana include: Ensuring high levels of employment Economic growth Price stability Interest rate stability Financial Market Stability and Stability of the foreign exchange rate.

1.6

Research Questions

This paper does not test for market efficiency, but tries to establish short and long run relationships between various macroeconomic factors and stock returns using the Ghana Stock Index for stock market performance. Macroeconomic instability had negative effect on the Ghana Stock Exchange. It is not clear from available statistics and earlier studies on the Ghana Stock Exchange what macroeconomic factors or combinations of factors are responsible for the performance of the exchange. In 1993 for instance, Inflation rose to 70% when,

21

the Interest rate was 30.95% and the Stock exchange recorded a performance rate of 116.06%. GDP grew in that year by 4.9%. In 1999, Inflation was as low as 21.3% and the Interest rate was as high as 34.19% but the stock performance was (15.14) % and GDP grew by 4.4%. The questions then are: 1: What macroeconomic factors drive the Ghana Stock Exchange? 2: How does GDP, Fiscal balance, Inflation rates, Interest rates and Exchange rates impact the Ghana Stock Index? Finding answers to these questions will help add to existing knowledge about the underlying causes of price movements of the Ghana Stock Index and how these variables can be used to predict market returns. It will be useful in giving policy guidelines to ensure stability of the capital market. It will also be a useful guide for Investors and Financial Analyst in assessing the price of stocks and their systematic risks with anticipated changes in the macroeconomic factors.

1.7

Interest in Research

This study is useful for Private Investors, Pension Funds, Government and policy makers because investments in equity have an assumption that corporate cash flows will grow with the economy, thus expected returns on equities may be linked to future economic performance. Macroeconomic factors impact both growth in the corporate sector and economy at large. Studying the relationship between macroeconomic factors and the stock market will assist in planning and predictability of both stock market and the economy. There is an emerging trend by which cohorts of investors lose their entire life time investment on stock markets or in Unit Trusts and Mutual Fund investments due to macroeconomic changes. Pensioners who expect lump sum payment from

22

Unit Trusts and other Investment Companies on retirement are shocked when their entire life time investments are wiped out or drastically reduced mainly because of the risks associated with stock markets. As indicated by David Swensen (2000), equity biased asset allocation yield higher returns but it involves the acceptance of higher risks. It is necessary for Trustees of Pension Funds and other Fund managers to assess the risks involved at each turn of macroeconomic variables in order to maintain risk and return profile of their funds. The Government of Ghana and the Central Bank are concerned about stability and growth. This study attempts to identify how sustainable stability can be obtained. The interest of this study is to identify the possible macroeconomic variables that impact the stock returns and to examine whether these macroeconomic factors are the cause of fluctuations in stock market performance. I also want to recommend appropriate strategies to ensure that macroeconomic and fiscal policies of Governments do not affect Investors, particularly Pensioners, considering the fact that Ghana has now introduced defined contribution schemes which is based on individual equivalence as against risk pooling. This study was undertaken in fulfilment of the requirements for the award of degree of Master of Science in Finance, University of Leicester. 1.7 Structure of Dissertation

The remaining chapters of this study are arranged in this order: Section 2, reviews literature and theory in order to identify the variables of interest to this study and to examine the results of other studies. Section 3, introduces the data, describes the data and explains the techniques used to examine and measure relationship between macroeconomic variables and movements in stock returns. It considers theoretical and empirical justification for modeling the Ghana Stock Exchange All Share Index as a proxy using specific macroeconomic variables, as well as the conceptual framework and

23

methodology. Analysis of Data was done in section 4. Section 5 discusses the findings and interpretation of results. Section 6 gives a summary. Section 7, gives the conclusions and section 8 considers recommendations and section 9 gives summary of reflections and experience gathered during the period of the dissertation. The last sections detailed the references and appendices.

24

2.0 2.1

LITERATURE REVIEW Macroeconomic factors and Stock Market Returns

According to Fama (1970), a stock market is efficient if current market prices fully and instantaneously reflect all available information about the macroeconomic fundamentals. Past information therefore contains no news so should have no effect on stock prices. Studies by Fama and Schwert (1977), Jaffe and Mandelker (1976) suggested that macroeconomic factors influence stock returns. Fama (1981), and Smith and Sims (1993) identified inflation, money supply, exchange rates as some of the major determinants of stock prices. Chen, Roll and Ross (1986) established the existence of long term equilibrium between stock prices and Inflation, Treasury–bill rate, Long term government bonds and Industrial production. They highlighted the fact that, economic factors affected expected dividends and discount rates. The ability of firms to generate cash flows and payout dividends is the basis of the long term equilibrium between stock returns and macroeconomic factors. Discount rates change with the level of interest rates, term structure and risk premium, expected dividends may change due to changes in inflation rate, production, and consumption levels. Fama (1981), Bodie (1976) and other writers made a strong case that inflation has a negative relationship with stock performance because high inflation rates add to uncertainty which reduces business confidence and thus lowers stock prices. This is in direct contradiction to the claim that stocks are hedge against long term inflation as indicated by Anari and Kolari (2002). Humpe and Macmillan reported that deflation has resulted in poor stock market performance of Japan. Interest rates change the discount rate in the valuation model and so influence current and future cash flows. Mukhererjee and Naka (1995) hypothesized that

25

changes in both short and long term interest rates will affect nominal risk free rates and so affect the discount rate. Fama and Schwert (1977) felt the relationship applies to both current period and for lagged observation of interest rates. Robert Pardy, (1992) emphasised the role of macroeconomic and fiscal environment in the development and performance of securities market. Corporate cash flows are related to a measure of aggregate output such as GDP or industrial production. Darrat and Mukherjee(1987), Mukherjee and Naka (1995) indicated that Stock Indexes are positively influenced by growth in GDP. Based on the fact that cash flows from firms are directly related to economic growth, Ritter (2004) argued that economic growth does not necessarily increase cash flow of existing stocks and stockholders. According to Ritter (2004) growth comes from high personal savings, increased labour force participation and technological change. If increases in capital and labour go into new corporations, it does to affect the cash flow of existing corporations. He continues that unless technological change comes from existing companies with monopoly power, it does not increase profits. It only increases per capita income of consumers. Many writers used Industrial production as a proxy of real economic activities. Fama (1990), Geske and Roll (1983) found positive relationship between Industrial Production and Stock returns through expected future cash flows. Some risks are involved in “dollarisation” of an economy as against the fact that investments are attractive when they are denominated in a stronger currency. Exchange rate appreciations are associated with higher investment leading to higher cash flows and higher stock performance. In an export oriented economy , as indicated by Mukherjee and Naka (1995) currency depreciations have positive relationship and impact the stock prices; deprecation of domestic currency leads to increase in demand of exports and cash flows. Currency appreciation, on the other hand, reduces competitiveness which results in a negative impact on the stock market.

26

Some studies considered relationship between the same macroeconomic factors and stock markets in different Countries. Funke and Matsuda (2006) considered USA and Germany. Humpe and Macmillan compared US and Japan. They employed the discount valuation model. They found contrasting results in each country. Other studies considered macroeconomic and stock market relationships in developed Countries like USA, Germany, UK and Japan. Flannery and Propapadakis (2002) examined developed countries like USA, U.K and Japan, whilst other writers like Choudhry (2001), Wongbagpo and Sharma (2002) considered developing Countries. Poon and Taylor (1991) studied the effect of macroeconomic factors on UK stock prices and found that macroeconomic variables do not affect stock market prices. 2.2 Models used to establish relationships between macroeconomic factors and stock returns Ross (1976) linked macroeconomic variables and stock market returns through an Arbitrage Pricing Theory (APT) using statistical tools like factor analysis. APT did not specify the factors but were statistically derived. The factors were fundamental economic aggregates like GDP, Inflation and Interest rates but these factors were not stated by APT. Multiple risk factors were used to explain asset returns. ATP concentrated on individual stock returns. It involves modeling a short run relationship between macroeconomic variables and stock prices in terms of first differentials, with the assumption of stationarity of the underlying data. Relevant studies which used this approach include; Fama (1981), Fama and French (1989), Schwert (1990) among others. Chen, Roll and Ross (1986) established the existence of long term equilibrium using specific macroeconomic variables, from the perspective of efficient market

27

theory and rational expectation and inter-temporal asset pricing theory. The macroeconomic factors were stated. The macroeconomic factors considered were Inflation, Treasury–bill rate, Long term government bonds and Industrial production. The Discounted Cash flow model was also used to establish relationship between macroeconomic factors and stock returns. This model links stock prices to future expected cash flows and the future discount rate of these cash flows. All macroeconomic factors that influence future expected cash flows and discount rates have an influence on the stock price. This model was also used to establish long run relationships between stock prices and macroeconomic variables. Campbell and Shiller (1988) established that Price Earning ratio predict stock returns over long period of time. They used earnings and expected dividends to establish the relationship between these factors and stock prices. The limitation of this approach is that there is no theoretical basis for selection of the variables. Jiranyakul (2008) used Present Value Model on the Thailand Stock market from April 1975 to December 2007 to evaluate whether the current market price of each stock deviates from its intrinsic value (present value of stock’s cash flows), from fundamental analysis. He concluded that there are other economic fundamentals other than dividends that cause stock price movements. Engel and Granger (1987), Granger (1986), Hendry (1986), Johansen and Juselius (1990), proposed the use of cointegration techniques and error correction models to establish long run and short run equilibrium between macroeconomic variables and stock market returns. Time series variables are cointegrated if they are integrated by the same order and a linear combination of all the variables is stationary. Linear combination of variables suggests the existence of long term relationship between the variables. The co-movement of the variables can be ascertained through error correction processes to establish short-term equilibrium. Cointegration and error correction models have been

28

used extensively to establish both long run and short term relationship between stock prices and macroeconomic variables in both developed and developing countries. This approach has become the preferred method of examining the relationship between stock returns and macroeconomic variables as indicated by Maysami, Howe and Hamzah (2004) and Utku Utkulu(). This approach was used by many writers including Mukherjee and Naka (1995), Maysami (2002) among others. The number of lags used in these models, particularly the error correction model, has a great impact on the results. 2.3 Previous Research on Ghana Stock Exchange

Twerefou and Nimo (2005) investigated the impact of asset pricing of the various sectors of the Ghana Stock Exchange for the period January 1997 to December 2002 using Arbitrage Pricing Theory. They concluded that inflation, short term interest rates and term structure of interest rates are macroeconomic factors affecting asset pricing in Ghana. Adam, Anokye and Tweneboah, (2008), examined long and short run relationship between interest rate, inflation rate, net foreign direct investment, and exchange rate and the Ghana stock market during the period January 1991 to December 2006, using cointegration and error correction models. They concluded that there is long run relationship between stock prices and the macroeconomic factors examined. They found out that there is a positive relationship between inflation and share prices. Kyereboah-Coleman and Agyire-Tettey (2008) used quarterly time series data of the following macroeconomic factors: Inflation, real exchange rates, Interest rates, and lending rates to examine the effect of these macroeconomic factors on the performance of the Ghana Stock Exchange. They concluded that lending rates have an adverse effect on stock performance and Inflation had a lagged

29

effect on stock performance. They were certain of the fact that depreciation favors investors. Frimpong (2009) used the cointegration model to establish long term relationship between exchange rates, the consumer price index, money supply, interest rates and stock returns of the Ghana Stock Exchange. He concluded that exchange rates have positive impact on the exchange while other variables have negative impact. These findings do not confirm each other mainly because of the lag period under consideration. Frimpong considered a lag period of nine (9) for cointegration analysis and lag of 3 for error correction. This study considers a lag period 10 to 15 for both cointegration and error correction.

2.4

Theoretical considerations

2.4.1 Interest rates Interest rates generally move in opposite direction with share prices. A fall in interest rates on money markets makes them less attractive in terms of returns. Investors generally react by transferring their investment to the stock market. This results in increase in demand for shares, this may lead to increases in prices. If interest rates increase on the other hand, investors may channel their current investments to the money markets thereby staving the stock exchange of the needed new investments. Trading activities therefore reduce as the market becomes bearish, as there are more shares on sale than buyers want, leading to fall in prices.

30

Theoretically, Interest rates are not very significant in the determination of firms’ investment decisions because most firms base their investment decisions on the profit expectation of investments. Investments, with returns exceeding costs (positive NPV) will be undertaken. Fluctuation in interest rates will lead to lower real returns, which will eventually prevent investment in some projects, resulting in reduced flow of investments. Most firms however, do not borrow to finance investments. They use internal funds or issue new shares.

2.4.2 Inflation In theory, stocks should be inflation neutral. This is based on the assumption that companies can pass on one-for-costs. Secondly, the required rates Investors used to discount cash flows do not rise when inflation rises. Moreover it is assumed that inflation does not have long term impact of growth. There is a theory to support the fact that Inflation is a hedge against long term inflation as indicated by Anari and Kolari (2002), because it represents claims against real assets and so stock returns should be positively related to expected inflation. Another theory supports the idea that Inflation negatively impact on stock prices. This is based on Fishers’ theory of interest which indicated that nominal interest rates may be decomposed into real rates and expected inflation. The argument here is that expected real returns are determined by real factors which are not related to inflation. 2.4.3 Fiscal deficits/Surpluses The main source of revenue for Governments is taxation. Personal and corporate taxes move in the same direction as growth of profits in the corporate sector. Fluctuations in government revenues are related to market movement. Governments do not limit their expenditure to revenue they spend more normally.. Deficits arise when expenditure is more than revenue and surpluses result when revenue is more than expenditure. Government treasuries borrow on 31

the open market to finance short fall in revenue. These debts are repaid in periods of surpluses. Government may also print more currency notes but this causes inflation. It can be said in general that, stock market returns have positive relation with Fiscal surpluses and negative relation with Fiscal deficits.

2.4.4 Exchange rates An exchange rate is the ratio of how many units of one currency you can buy per unit of another currency. Unanticipated currency movements results in risk of changes in the value of assets and liabilities of a firm. It impacts on sales, prices and profits of importers and exporters. It also reduces competitiveness. Purely domestic firms which are not involved in imports and exports may also suffer from exchange rate risk when they compete with foreign companies in their home markets. The three main types of foreign exchange risk exposures are translation, transaction and operating exposure. There are three factors which cause the currency of a Country to appreciate or depreciate. These are: Differences in income growth: Nations with high income growth will demand more imported goods all other things being equal resulting in demand for more foreign currency leading to appreciation in foreign currencies relative to domestic currency. Differences in inflation rates: Consumers in a country with higher inflation rate will demand more imported (cheaper) goods from other Countries leading to appreciation of foreign currencies and depreciation of local currency. Differences in real interest rates: Differences in real interest rates results in a movement of capital from countries with lower real interest rates to countries with higher interest rates.

32

2.4.5 Gross Domestic Product (GDP) Economic factors that impact on changing investment opportunities include macroeconomic factors which are measured by Gross Domestic Product (GDP), a measure of economic performance. GDP does not include purchase and sale of stocks since these are not production of goods. GDP measures both output and income. If output increases per person this translates into higher standard of living. GDP does not measure homemaker services and other non market production nor underground economy e.g. Illegal activities and tax evasion. Economic growth is the increase in value of goods and services

produced in an economy. It results from saving and use of capital, increase in work force and from technological changes. It is measured by the percentage increase in the real Gross Domestic Product (GDP). GDP is the total market value of all domestically produced final goods and services in a given year. It includes income earned by foreigners but excludes income earned by citizens abroad. The Expenditure approach is measured by the summation of the following. GDP = C + I + G + X, where (C) is household consumption of goods and services. ( I ) is expenditure of businesses. It includes capital replacement and new additions to capital assets, as well as investments in Inventory. (G) is Government consumption and gross investment. It includes purchases of goods and services by Government and its agencies. It excludes transfer payments. (X) is net export of goods and services.

33

Current year values are used with a GDP deflator and base year values to derive real GDP. Factors that affect aggregate demand (GDP) in an economy include: Real Wealth: An increase in consumers’ real wealth will increase aggregate demand because of an increase in consumption. A fall in real wealth will lead to fall in consumption and for that matter GDP decreases. Real Interest Rates: An increase in real interest rates will lead to a decrease in consumer expenditure and business investment resulting in a decrease in aggregate demand. A decrease in real interest rate on the other hand increases aggregate demand through reduction in finance costs for both businesses and consumers. Inflation expectation: An expectation of future increase in inflation will increase aggregate demand whilst a decrease in the expected rate of inflation will reduce current purchases and aggregate demand. Exchange rate fluctuations: Currency appreciation leads to an increase in the value of the domestic currency (GH¢) compared to other currencies. This makes imports less expensive leading to an increase. Exports on the other hand become more expensive to foreign consumers leading to a fall. Since Ghana imports more than it exports on aggregate basis, Net Exports (Exports-Imports) will be negative leading to a fall in aggregate demand. Currency depreciation will on the other hand increase exports and reduce imports leading to increase in aggregate demand. Economic Expectations: Positive expectation about the future prospects of the economy increases aggregate demand resulting

34

from increase business investment in capital in anticipation of increased future sales. Higher productivity from both capital and labour will increase aggregate demand. Investment in research and development as well as Innovations and technological development will increase aggregate demand. The rate of local capital accumulation also affects aggregate demand. Government policies in encouraging education, technological development, trade promotion, low taxes and high savings rates. An economy with increasing GDP all other things being equal will have rising Income. This will lead to an improvement in the disposable incomes of individuals. Demand for shares which is related to the level of disposable income will increase leading investors to buy more shares; the higher the demand for shares and higher the share prices to move up. When the economy is sluggish, the level of income and hence disposable income is affected negatively. There is no increase in production, employment is low leading to loss of jobs and reduced disposable income for most workers resulting in investors cutting back on their investments in shares. The resulting fall in demand for shares leads to a fall in share prices. 2.5 The Role Governments

Government can improve economic growth by adopting pragmatic policies that encourage growth by constantly monitoring the business cycle (contraction, recession, expansion and business peak). Governments can create an enabling environment by providing infrastructure (educational, technological, financial, physical, environmental, and social) as well as tax regimes and general confidence in the business environment.

35

Governments as agents of development and growth can also use their consumption and Investment to facilitate growth. Replacement, new and capacity enhancement investment in facilities that boost business production will provide good investment climate. Government’s investment in infrastructure is a necessary tool in boosting business confidence.

Governments wishing to promote economic growth may endeavour to maintain incremental macroeconomic stability. They should avoid budget deficits and excessive surpluses. Time lags in recognition, implementation and impact of policy should also be monitored and controlled to ensure policy interventions yield desired results.

36

3:0

Methodology

This section provides an overview of the data collection, variable selection, model specification and tests conducted. There exists no theoretical framework for selection of macroeconomic variables that affect stock market performance. Economic theory suggests that stock market prices should reflect expectation about future corporate cash flows which generally reflect the level of economic activities relate to aggregate output such as Industrial Production or GDP. If stock prices reflect the underlying fundamentals, then the stock prices could be used as indicators for future economic activities. It can be said therefore that the relationship between stock prices and macroeconomic variables are useful in formulation of macroeconomic policies. 3.1 Data Collection

Monthly time series data from January 1991 to December 2008 was used in the model. Data on Fiscal balance, Inflation, 91 day Treasury bill rates and Exchange rates movements were obtained from the Bank of Ghana. Stock Index movements were obtained from the Ghana Stock Exchange. The Ghana Stock Exchange, All Share Index (GSI) is a composite index which measures price movements of all equities listed on the exchange. The index is based on the low and high prices within the month/year as against beginning and closing figures. 3.2 Variable Selection

There is no economic theory that explains the linkage between macroeconomic factors and stock market performance but there are several macroeconomic factors that have been identified as having impact on stock prices. This study uses dividend discount model to select the variables considered in the study.

37

Returns on stocks market are two fold. Capital Gains (price changes) and Dividend yield. Capital gains are the percentage increase or decrease in the price of an investment and includes gains or losses due to changes in exchange rates. Capital gains are computed by (P1/P0-1)*100. Dividend yield is stated as a percentage of price (D1/P0 )*100. Where P1 is current period prices and P0 is last period prices and D1 is current period dividend per share. This study considers only capital gains computed on the basis of high and low price movements in the month/year. Dividend yield is excluded because of lack of data. Changes in the earning capacity of a firm depend on general economic and market conditions. A change in individual stock returns depends on changes in rates of return for the entire stock market and the stock’s Industry. This requires examination of Macroeconomic, Industry and Individual stock performance, but this study is limited to the impact of macroeconomic factors on the entire stock market irrespective industry differences. Gordon’s Dividend Growth model assumes constant growth. The value of a stock is derived from P0 = . D0(1+g) k-g or D1 k-g

where, g is constant growth in dividend, k is the required rate of return and D1 is expected dividend, which is a product of current dividend and growth rate. It is clear from the equation above that price depends on market discount rates and expected stream of dividend payments and growth factors. Expected inflation is built up in prices and projections so it may not affect stock prices, but unanticipated inflation may directly influence stock prices through changes in price level and through increases in discount rates. An increase in

38

discount rates reduces, the present value of corporate cash flows. The impact of inflation on discount rates and price levels makes it imperative for its inclusion in model. Money supply directly impact on interest rates. Interest rates influence investment decision on holding non interest bearing securities or interest bearing securities. At high interest rates, Investors may sell Equities to invest in Treasury Instruments. Interest rates change the discount rate in the valuation model and influences current and future cash flows. Long term bond yield will have been ideal for this studies, due to lack of data availability we use the 91 day treasury bills, which is more sensitive to the market at all times than the one year rate in an inverted term structure of interest rates which currently exists in Ghana. The Ghanaian economy depends largely on loans, donor support and remittances from citizens of Ghana resident abroad. The shares of AGC and Golden Resources are traded on foreign stock markets. Prices of houses, cars and other consumer items are quoted in dollars. Fluctuations in the dollar rate, which represent all foreign currencies, have significant impact on the economy therefore its inclusion in the model.

3.3

Model Specification

The GSI is the dependent variable of the regression equation. It represents the performance indicator of the Ghana Stock Exchange. Fiscal Deficits/Surpluses captures the income and expenditure position of the economy; Interest rates are approximated by 91 day Treasury bills; Inflation rate and Exchange rates are the other independent variables. The model is:

39

GSI = β0 + β1FD + β2INTrate + β3INFrate + β4EXrate + β5GDP + et

Eq1

where GSI is the Ghana Stock Index, FD is Fiscal position, INTrate is interest rate, INFrate is inflation rate, EXrate is the exchange rate and GDP is gross domestic product, β0 - β5 are coefficients of variables and et is the error term, representing others factors not considered in the model e.g. Oil price hike and change in Government. To examine causal relationship of the variables we specify the following multivariate model: U = (GSI, GDP, Fiscal, Interest, Inflation, Exchange) where GSI is the Ghana Stock Index, GDP is gross domestic product. Fiscal is the fiscal position of Income and Expenditures, Interest is interest rate, Inflation is inflation rate, and Exchange is the exchange rate. 3.4 .0 Tests conducted 3.4.1 Visual Inspection: Visual inspection of the time series data was observed and comments made on movement in the variables. 3.4.2 Unit roots Tests Standard inference procedures do not apply when regressions contain integrated dependent variables or integrated regressors so it is necessary to ascertain whether a series is stationary or not before using regressors since the eventual results may be spurious. A series is stationery if the mean and auto-covariance are not time dependent. If the first differences are stationery, the series is said to

40

be integrated by the order 1(d) where d is the order of integration. The order of integration defines the number of unit roots contained in the series or the number of differentiations it takes to make the series stationary. Stationarity ensures that the variables are stationary and that shocks are temporary and will revert to long term mean after the effect of the shock. It is useful in deriving meaningful statistics such as means, variances and correlations. Unit roots tests was applied using Augmented Dickey Fuller (ADF) test with automatic maximum lag of 14 and Schwarz Information criterion to ascertain the order of integration of each variable to ensure stationarity of the variables. The automatic lag period 14 measures the correlation between observation 1 and 15, observation 2 and 16 in that order. The existence of unit root means the data set is non-stationary. Nonstationary time series data tend to be auto correlated whilst stationary time series data tend to be random. The ADF calculates a test statistics which if greater than the Dickey – Fuller critical values, the null hypothesis can be rejected and conclusions drawn that the data had unit roots and so they are stationary. The critical values used for this study is -2.875 at 5%. We first test the original data. It was found to be non stationary so we make the data set stationary by taking the first difference transformation, in order to eliminate the trend relationships in the data. 3.4.3 Cointegration Tests To establish long term relationships and co-movement among the variables, Co integration techniques by Johansen and Juselius (1990), Johansen (1991) protocols was used instead of Engle and Granger (1987) because the analytical software EViews is limited to Johansen protocol. The Johansen’s (1991) Vector Error Correction Model also allows testing for cointegration in a whole series of equations in one step without requiring normalization of variables unlike the two step approach of Engel and Granger (1997).

41

If two series are co integrated or have long term relationship, it means there exists a stationery linear combination of these series and the series are of the same order. A unit root test for stationarity of the time series data is first determined before the cointegration test. This test determines the existence of a unit root test of each series. The series are observed to determine whether they are stationary or they are integrated of the same order. If two variables are nonstationary but stationary in first difference, the series can be said to be integrated of order one. Thus they are I(1) series. Cointegration tests were performed using Johansen VECM using ordinary least square regression. There are two tests; the trace statistics which tests the null hypothesis that there are at most r cointegrating relationships and the eigenvalue statistic which tests the hypothesis of r cointegrating relationships against the defined alternative of r+1 cointegrating relationships. examined. 3.4.4 Vector Error Correction Cointegration was confirmed so Johansen and Juselius (1990) Error Correction Model was used to establish short run causal relationship between the two variables. Vector Error Correction model is a system of equations for which, each variable is a function of its own lag and the lag of other variables in the system. The variables have an order 1 and are cointegrated. The F tests of the explanatory variables determined indicate short run relationships. 3.4.5 Granger causality tests Granger causality tests using bivariate vector autoregressive method was used for all pairs of the series in the group. It measures the precedence and information content but does not by itself indicate causality in the more common An advantage of the cointegration analysis is that co-movement among the variables can be

42

use of the term. It examines short term causal relationships between the stock returns and each variable by calculating F statistic of the joint hypothesis. If one variable does not improve the forecasting ability of the other variable, then the first variable does not Granger cause the second. If the F statistic is significant, we can reject the null hypothesis that variable 1 does not Granger cause variable 2. 3.5 Data Description

The data covered a period of eighteen (18) years with two hundred and sixteen (216) observations on GSE, real GDP, FISCAL, INFLATION, INTEREST and percentage change in Exchange Rates. GES, real GDP, INFLATION and INTERST rates are expressed in percentages. Fiscal is the fiscal balance as a percentage of GDP Table below gives the summary of the data.

Table 2. Summary of Data Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque-Bera Probability Sum Sum Sq. Dev. Observations GSE 43.75667 30.84500 154.6700 -15.22000 47.93129 0.870153 2.699739 28.06942 0.000001 9451.440 493942.8 216 GDP 4.883333 4.650000 7.300000 3.300000 0.994637 0.657757 2.952106 15.59582 0.000411 1054.800 212.7000 216 FISCAL -4.522222 -4.580000 2.260000 -11.48000 3.658188 0.200122 2.356296 5.170946 0.075360 -976.8000 2877.203 216 INFLATION 24.08683 18.11500 70.80000 0.390368 16.71154 1.466142 4.630702 101.3173 0.000000 5202.756 60044.22 216 INTEREST 28.22486 27.00000 47.93000 9.600000 12.01295 0.119472 1.922574 10.96147 0.004166 6096.570 31026.87 216 EXCHANGE -1.172062 -0.935779 0.181738 -3.365348 1.126772 -0.526171 1.876563 21.32581 0.000023 -253.1654 272.9673 216

The sample period recorded average fiscal deficit of 4.5% and exchange rates show an average percentage change of 1.17% change. The Ghana Stock Exchange shows a mean performance rate of 43%. All variables are positively skewed above normal with the exception of exchange which is negatively skewed. It indicated that there are outliers which are greater than the mean in all

43

the variables with the exception of exchange which have outliers less than the mean. All the variables have relatively low Kurtosis. They are all more or less peaked than a normal distribution which has a Kurtosis of 3 standard deviations. GDP looks normal whilst GSE, Fiscal, Interest and Exchange are less peaked than normal distribution. Inflation is more peaked than normal distribution. Jarque- Bera statistics significantly rejects normal distribution of all the variables indicating non-normality. The coefficient of variation (Standard Deviation/Mean) as shown below is used to examine the volatility of each macroeconomic factor. GSE GDP FISCAL -0.81 INFLATION INTEREST 0.694 0.425 EXCHANGE -0.96

1.095 0.20

This shows that the Ghana Stock returns are very volatile, recording over 100%. Exchange rate, and Inflation and Fiscal fluctuations also show wide variability. GDP and Treasury bill rates are more stable than the other variables. 3.6 Regression

Table 3 Regression

Dependent Variable: GSE Method: Least Squares Date: 11/23/09 Time: 10:13 Sample: 1991M01 2008M12 Included observations: 216 Variable GDP FISCAL INTEREST INFLATION EXCHANGE R-squared Adjusted R-squared S.E. of regression Sum squared resid Coefficient 6.408043 1.376725 -0.043016 0.992898 -2.081595 0.148491 0.132349 44.64693 420596.5 Std. Error 1.563947 0.868126 0.226801 0.199781 1.291448 t-Statistic 4.097352 1.585859 -0.189666 4.969923 -1.611831 Prob. 0.0001 0.1143 0.8498 0.0000 0.1085 43.75667 47.93129 10.45832 10.53646

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion

44

Log likelihood Durbin-Watson stat

-1124.499 0.134767

Hannan-Quinn criter.

10.48989

Table 3 shows the results of multiple regressions of all the variables. The results indicate that less than 15% of movement in the stock market is explained by the independent variables. This regression is not valid because the variables are cointegrated. Multiple regression analysis alone is not be able to answer the research questions since the variables of interest are interdependent and the analysis require a test of the stationarity of the data to establish the causality of price movement and dependencies of the variables. The original time series data was used to conduct an Ordinary Least Squares regression between two variables at a time under the model: Yt = a + Bxt + ut

Eq2

Where, ut is the estimated residuals on the long run equilibrium to establish the existence of unit roots.

3.7.0 Hypothesis The following hypotheses were made about the relationships between short and long term interest rates, inflation, exchange rates, fiscal balances and GDP. 3.7.1 Interest rates Dividend discount valuation model is very elaborate on the impact of interest rates on the price of stock. In Gordon’s constant dividend model,

45

P =D1/ (k-g) ------------------------

equation 1,

Where P is the stock price, D1 is dividend after first year, g is the constant growth in dividend, and k is required rate of return. Changes in both short and long term interest rates affect discount rates used by investors to evaluate projects. From the equation above, if k is greater than g, there is a negative relationship between k and P. If k increases, P reduces. The first hypothesis is that: There is a negative relationship between interest rates and stock prices.

Intuitively, interest rates influence corporate profits, and therefore expected future dividend payments. The lower the interest rates the more profit available for reinvestment or for distribution, the more future expectation and the more willing investors are to buy the shares leading to increase in prices. Investors using borrowed funds will enjoy reduction in interest rates because of its impact on cost of borrowing. Reduction in interest rates will increase demand for shares since investors will require lower rate of returns to buy shares. High interest rates encourages investors to make quick money by buying treasury bills at the expense of stocks this results in reduced demand and eventually drive down prices. 3.7.2 Inflation The real rate of interest is the nominal rate minus inflation. Both short and long term inflation will directly result in higher real rates leading to higher discount rates. From equation 1, the hypothesis stands as above for interest rates that:

46

There is a negative relationship between inflation and stock prices on condition that inflationary movements are unanticipated. High Inflation rates increase the cost of living and divert resources from investment on the stock market to consumables and real estate among others whose prices are on the increase in order to preserve capital. This results in low trading volumes and general lack of liquidity of the stock market. Lack of liquidity and low demand constrain traders to accept discounts on shares they offer for sale, this culminates in low prices. Inflation may lead to increase in cash flows but the increase can not buy the same basket of goods and services because cash flows will not grow at the same rate. An increase in Inflation directly increases the nominal risk free rate of interest which raises the discount rate resulting on negative impact on the price of shares and the general performance of the Stock Index. If inflation is expected however, all pricing in the economy will include inflation expectation, stock market prices are relatively well priced, so one will expect that inflation will have positive effect on the stock market returns.

3.7.3 Exchange Rates The Ghanaian economy is import dependent. Trade balances are mostly in deficit. Depreciation of the currency against the dollar in particular hurts the economy greatly, as cost of production goes up, corporate profits reduce. Depreciation of the cedi however favours exporters. Export oriented companies on the exchange will experienced increase economic activity with depreciation but since Ghana is import dependent, I hypothesis that:

47

Depreciation of the Ghanaian currency has negative impact on the stock exchange and stock prices. 3.7.4 Fiscal Deficits/Surpluses

An increase in budged deficit requires more money to finance the deficit resulting in increased money supply or open market operation by the Central Bank. Increased interest rates which come with selling Treasury instruments stave listed companies of needed capital as returns on treasury instruments become very lucrative far above the short term returns of the stock market. I hypothesis that: There is a negative/ (positive) relationship between budget deficits/ (surpluses) and stock market returns. 3.7.5 Gross Domestic Product

Gross Domestic Product measures the real economic activities. It is high during periods of economic growth and low during periods of contraction. I hypothesis that; There is a positive relationship between GDP growth and stock market returns. 4.0 4.1 Data Analysis Unit Root Tests

It is necessary to ascertain whether a time series data is stationary or nonstationary before relying on regression analysis because there is a danger of

48

obtaining regression results from unrelated data from nonstationary series yielding spurious regression. Unit Roots tests for stationarity were used to identify the order of integration. The results listed below shows that the calculated Augmented Dickey – Fuller (ADF) test is greater than the test statistics so we do not reject the hull hypothesis of nonstationarity, thus all the variables are nonstationary. We however need to transform the series to stationary. The lag length used for the ADF test is based on Schwartz information Criterion (SIC). The indicative lag ranged from zero (0) to twelve (12).

**Table 4: Summary of Level 1 ADF Unit Root Test *
**

Variables GSE GDP FISCAL INTEREST INFLATION EXCHANGE Levels -2.705 0.265 -1.403 -1.619 -2.282 0.045 t-statistic -2.875 -2.875 -2.875 -2.875 -2.875 -2.875 Lag Length 0 12 0 1 1 3

* Calculated ADF is greater than 5% critical value so we do not reject the null hypothesis of non stationarity

The original series based on the ADF tests are non stationary so we take the first difference transformation of the non stationary data and based on ADF test. The series is now stationary because the calculated ADF is lower than the test statistic as shown in Table 5. We therefore reject the null hypothesis of nonstationarity.

Table 5: Summary of Level 2: ADF Unit Root Test **

49

Variables ΔGSE ΔGDP ΔFISCAL ΔINFLATION ΔINTEREST ΔEXCHANGE

First Difference -14.595 -8.031 -14.594 -10.459 -12.595 -12.876

t-statistic -2.875 -2.875 -2.875 -2.875 -2.875 -2.875

Lag Length 0 11 0 0 0 2

** Calculated ADF is less than 5% critical value so we reject the null hypothesis of non stationarity.

This shows that an equilibrium relationship has been established to enable us deduce long run relationships. This implies two or more variables cannot move independently of each other and data from a linear combination of any two variables can be stationary despite the fact that those variables may be individually non stationary, this is however only possible if the variables are integrated to the same order (order 1) as the trends in both series cancel each other. 4.2 Cointegration

Methodology of Johansen (1991) was used to test the model to determine the rank, r and to find the cointegrating relationships in the model. Selection was made for intercept and “no trend” for the cointegrating equation. It was noted that the lag length was very relevant in the long term relationship of the variables, lag length of 2 to 6 for instance yielded no cointegrating equation but lag period of 10 to 15 showed one cointegrating equation for both Trace and Eigenvalue tests at 5% significance level, as indicated in Appendix B. The null hypothesis for the Trace test, is that the number of cointegrating vectors is less than or equal to 1 against an alternative hypothesis that there are more than 1 cointegrating vectors. The Maximal Eigenvalue test has the null

50

hypothesis is that there are r cointegrating vectors present against the alternative that there are (r+1) present. The normalized cointegrating coefficients for the model is shown below Yt = (GSEt β1 = (1.00 GDPt, - 4.19 FISCALt + 3.12 INTERESTt + 2.14 INFLATIONt - 3.81 + EXCHANGEt) is 1.21)

**The cointegrating relationship can be re expressed as
**

GSE = 4.19 GDP - 3.12 FISCAL – 2.14 INTEREST + 3.81 INFLATION – 1.21 EXCHANGE

(3.8)

(7.2)

(2.9)

(1.8)

(9.0)

The coefficients are all significant and those of GDP and Inflation are positive whilst those of Fiscal, Interest and Exchange rates are negative. This indicates that GDP and Inflation are positively related to the GSE and negatively related to Fiscal, balances, Interest rates and Exchange. The existence of cointegration implies significant error correction term, which also is an indirect test of short run causality. The coefficients are significant. It indicates co-movement between stock market performance and the macroeconomic variables of interest in this study. It establishes long run equilibrium relationship.

The table 7 below shows the normalized co-integration equation and the adjustment coefficients. Table 7 Normalized Co-integration equation

1 Cointegrating Equation(s): Log likelihood 2103.563

51

Normalized cointegrating coefficients (standard error in parentheses) GSE GDP FISCAL INTEREST INFLATION 1.000000 -4.19E-16 3.12E-16 2.14E-16 -3.81E-18 (3.8E-08) (7.2E-09) (2.9E-09) (1.8E-09) Adjustment coefficients (standard error in parentheses) D(GSE) -1.000000 (1.6E-09) D(GDP) -0.002605 (0.00116) D(FISCAL) 0.003518 (0.00528) D(INTEREST) -0.002204 (0.01025) D(INFLATION) -0.037004 (0.02077) D(EXCHANGE) -3.82E-05 (0.00041)

EXCHANGE 1.21E-14 (9.0E-08)

The movement away from the long term equilibrium is determined by the adjusting coefficients. This shows insignificant movement with the exception of Exchange which responds quickly to short term shocks in the opposite direction. The coefficient of -1.2 compared with adjusting coefficient of 3.8, is possible to affect the long run equilibrium.

4.3

Vector Error Correction Analysis

Since the variables have an order 1 and are cointegrated, we proceed to examine the short run causal relationship between the Ghana Stock Index and the macroeconomic factors of interest, using the Vector Error Correction model, which is a system of equations for which each variable is a function of its own lag and the lag of other variables in the system. It is useful for identifying and testing the short run equilibrium relationship among the variables and to ascertain the impact of each macroeconomic variable on the stock market returns. The error correction term and their joint significance are provided by the F statistics. The F statistics listed in Appendix 3 range from 1.59 to 5.96. This rejects the null hypothesis that the coefficients of the variables are equal to zero. As shown in

52

the error correction equation in table 9 below: all the variables contribute towards the error correction process. This indicated short term relationship. Table 8 Error Correction equation

Error Correction: CointEq1 D(GSE) -1.000000 (1.6E-09) [-6.2e+08] D(GDP) -0.002605 (0.00116) [-2.23989] D(FISCAL) 0.003518 (0.00528) [ 0.66682] D(INTEREST) D(INFLATION) D(EXCHANGE) -0.002204 (0.01025) [-0.21502] -0.037004 (0.02077) [-1.78167] -3.82E-05 (0.00041) [-0.09390]

The table above shows that Interest and Exchange have coefficients in opposite direction but all the other variables have error correction coefficients with the right sign. The speed of adjustment appears to be fastest for Exchange but it not statistically significant. The other variables are very slow this gives an indication that it takes a long time for stock market to get back to equilibrium after changes in macroeconomic variables. This result suggests less effective short term relationship between the Ghana Stock Market returns and the macroeconomic variables.

**Table 9 Regression of Stationary Data
**

D(GSE) D(GDP) D(FISCAL) D(INTEREST) D(INFLATION) D(EXCHANGE)

R-squared Adj. R-squared Sum sq. resides S.E. equation F-statistic Log likelihood Akaike AIC Schwarz SC Mean dependent S.D. dependent ,

1.000000 1.000000 5.90E-25 6.05E-14 4.34E+29

0.308450 17.43455

0.431654 0.297510 5.832479 0.190333 3.217848 69.69982 -0.306998 0.336174 0.017000 0.227088 1.07E-29 2.92E-30 2038.733 -17.98733 -14.02935

0.355578 0.203478 119.9482 0.863145 2.337793 -232.6620 2.716620 3.359792 -0.048500 0.967131

0.272909 0.101298 452.8057 1.677038 1.590273 -365.5023 4.045023 4.688195 -0.003650 1.769030

0.558877 0.454761 1859.414 3.398405 5.367835 -506.7576 5.457576 6.100748 -0.080350 4.602373

0.584641 0.486606 0.712664 0.066532 5.963584 279.9186 -2.409186 -1.766014 0.005793 0.092855

Determinant resid covariance (dof adj.) Determinant resid covariance Log likelihood Akaike information criterion Schwarz criterion

53

Table 10 shows regression of stationary data. R2 show low to medium explanatory power of the variables considered. 4.4 Granger Causality

The existence of relationship between macroeconomic factors examined above does not help in determining the direction of causality. We select a lag length of fifteen (15) to conduct Granger causality tests to examine the causal relationships of each variable and the stock market returns using bivariate vector autoregressive model. explanatory variable. F test was used to test for the significance of each

From the table above, it can be concluded that: Fiscal Granger cause GSE and GSE granger cause Fiscal in a bivariate direction. Exchange Granger cause GSE but GSE does not Granger cause exchange. It is unilateral relationship. GDP Granger cause Exchange but Exchange does not Granger because Exchange It is a unilateral relationship Fiscal Granger cause inflation but Inflation does not Granger cause Fiscal. It is a unilateral relationship Exchange Granger cause Inflation and Inflation Granger cause Exchange, in a bivariate relationship. 5.0 Research Findings

54

Results obtained from the analysis of data indicated that macroeconomic factors considered in this study have unit roots and so multiple regression analysis will not yield any meaningful results. Cointegration analysis shows that all the variables of interest have significant coefficients indicating long term relationship. The speed of correction in the short term is slow; this is confirmed by Granger causality test. The results indicate that the Ghana Stock exchange has a positive long run relations ship with GDP and Inflation but negative long term with fiscal, interest rates and exchange rates. These results are substantially in line with findings reported in earlier studies of the GSE. Anthony Kyereboah-Coleman (2008), Kwame Agire-Tettey (2008) and Frimpong (2009) reported that Treasury Bill rates (Interest) have negative impact on Ghana Stock Exchange returns. High interest rates will lead to diversion of investment funds from the exchange to money market. This study indicated that Inflation has a positive influence on the Ghana Stock Market returns. This confirms the findings of Anthony Kyereboah-Coleman (2008), but is contrary to the findings of Kwame Agire-Tettey (2008) and Frimpong (2009). This study also indicate exchange rates have negative influence on the market which confirms the findings of Anthony Kyereboah-Coleman (2008), but is contrary to the findings of Kwame Agire-Tettey (2008) and Frimpong (2009) as in the case of Inflation. The comparative results are shown in table 12. Table 12 Comparative results of studies on GSE

Macroeconomic Factor

GDP FISCAL INTEREST INFLATION EXCHANGE

Coleman& Tettey N/A N/A +

Anokye & Tweneboah

N/A N/A + -

Frimpong N/A N/A +

This Study + + -

The differences in findings can be explained firstly by consideration of expectation theory and the lag period considered by the individual studies.

55

Inflation expectation makes Companies build their expectation in their prices for both long run and short run this deviated from the theoretical expected position. This study confirms the fact that the Ghana Stock Exchange does not respond quickly to changes in macroeconomic factors. Anokye and Tweneboah (2008) reported that the Exchange responds less to real activity than to monetary shocks. From their studies, it took five (5) quarters or twenty (20) months for shocks in Treasury Bills rates to have an impact on the Stock market. The period range from five (5) quarters to twelve (12) quarters which is essentially medium to long run. The transmission mechanism of macroeconomic factors need further studies to understand why changes in macroeconomic factors take a long time to have an impact on the market. A weak currency increase the price of imported goods so Companies of a net importing Country like Ghana will suffer with currency depreciation. Findings suggest that the stock exchange and the cedi move in the opposite direction, appreciation in the cedi reduces stock market performance and depreciation of the cedi strengthens stock market performance. This analysis emphasis the importance of lag in the implementation of economic policy based on expectation of economic activity. The lags include recognition, implementation and impact. The impact lag is shorter for fiscal policy than monetary policy. Unanticipated changes have greater impact than anticipated policy changes. Maintaining fiscal discipline was and is still a major challenge for Governments of Ghana. Stability can be achieve through conscious effort towards: Stable monetary growth, managing money supply with the hope of maintaining prices; maintaining fiscal deficit/ (surplus) during recession/ (expansion) by establishing automatic stabilizers.

These findings differ from the hypothesized expectations only in the case of inflation. Expected inflation have a positive relation with the GSE but it was

56

expected that unanticipated inflation will have negative impact on the exchange but results of this study contradict that expectation. The justification could be that the pricing mechanism is very effective in inflation forecasting. Depreciation of the currency have negative impact on the GSE All Share Index but the short run adjusting coefficient suggests that depreciation of the Ghana Cedi could have positive relation with the market.

6.0 1.

Conclusions This paper examined the long and short run relationships between five (5) macroeconomic variables and Ghana Stock Exchange All Share Index for the period of eighteen (18) years, from January 1991 to December 2008. The macroeconomic factors considered were Gross Domestic Product, Fiscal balance, Inflation, Interest Rates and Exchange rate.

2.

There is no economic theory that explains the linkage between macroeconomic variables and stock market performance in any one direction but there are several macroeconomic factors that have been identified as having an impact on stock prices. This study used dividend discount model to select the variables used in the study.

3

The existence of cointegration between the variables was tested by establishing the order of cointegration, using unit roots test and then by performing rank tests using Johansen procedure to establish long run relationships before error correction and granger causality test were used to establish short term relation ships.

4

The effects of macroeconomic factors on the performance of stock markets cannot be generalized in a set in hypothesis or theories that is applicable to all nations. There are country differences depending on the

57

structure of the economy and the policy transmission mechanism. Stock returns are normally influences by both international and local conditions. 5 The results of cointegration indicated single cointegration vector between the GSI and the macroeconomic factors examined. The coefficients from the cointegration vector normalized on GSE provided evidence of long run relationship between the GSI and the economic variables considered. The results provided evidence that the Ghana Stock Exchange is positively related to GDP and Inflation, and negatively related to Interest rates, Fiscal movements and Exchange rate fluctuations. 6 Empirical evidence in this study support expected relationship between the performance of the GSI and the macroeconomic variables considered, with the exception of Inflation. This possibly reflects the fact that Inflation expectation is factored in the pricing mechanism of the economy so only unanticipated inflationary movement can have a negative impact on the Ghana Stock Market and the economy at large. This suggests that the Bank of Ghana’s Inflation targeting regime may be working well. It is also possible that the use of nominal rates instead of real rates might have contributed to this result. Depreciation of the currency have negative impact on the GSE All Share Index but the short run adjusting coefficient suggests that depreciation of the Ghana Cedi could have positive relation with the Exchange. The pace of adjustment is however slow. 7 The role of Government in crafting and implementing fiscal and monetary policies have significant impact on the economy and therefore the Capital Market but there are other factors like oil price hikes; change of Government; international financial and economic development etc. which have major impact on the economy of Ghana, but these do not form part of this study.

58

8

Ghana has a well though out fiscal and monetary policy but the policy transmission mechanism has some challenges, largely because of implementation of economic policy without careful consideration of the period of lags on macroeconomic factors. Maintaining fiscal discipline was and is still a major challenge for past and current Governments of Ghana. Governments wishing to promote economic growth may endeavour to maintain incremental macroeconomic stability. They should avoid budget deficits and excessive surpluses. Time lags in recognition, implementation and impact of policy should also be monitored and controlled to ensure policy interventions yield desired results.

9

Ghanaian policy makers need to be careful when trying to influence the economy through in macroeconomic factors. The oil fine for instance is expected to lead to increased supply of foreign currency, this may be channeled into to economy to solve problems of unemployment, and this may lead to excessive appreciation of the currency making export from other sectors less competitive. Increased money supply may also lead to inflation.

10

This study emphasise the importance of lag in the implementation of economic policy. The lags include recognition, implementation and impact. The impact lag is shorter for fiscal policy than monetary policy. Unanticipated changes have greater impact than anticipated policy changes.

11 12

The Government of Ghana is concerned about stability and growth. The lag length used for the error correction in this study may be a limitation and may account for reasons of disparity in results when this study is compared with other studies conducted on the GSI.

59

8.0

Recommendations

1. The Government of Ghana is concerned about stability and growth. Greater stability can be achieved if the Government of Ghana establishes the following through the Bank of Ghana and the Ministry of Finance. Stable monetary growth to match the future growth potential of the economy. Manage money supply with the hope of maintaining prices; Maintain fiscal deficit or surpluses during recession and periods of expansion respectively; Use automatic stabilizers like unemployment benefits during periods of low economic activities, high corporate taxes during booms and low corporate taxes during recession, and wider coverage and progressive personal income tax regimes. Use of Oil revenues with caution to avoid negative impact on macroeconomic variables. 2. The policy transmission mechanism needs to be studied thoroughly to understand why it takes so long a time for the impact of policy to be felt in both the Economy and the Stock Market. The lags in policy identification, implementation and impact should be studied and carefully eliminated to ensure intended actions do not have lagged effects.

60

3. Future studies of Ghana Stock Exchange and macroeconomic factors should use real rate of returns instead of nominal rates.

8.0

Reflections

Research work on this topic began when I met Dr. Tomasz Wisniewski at the 2008 Summer School. I went to the Summer School at the time I was considering my dissertation topic. I wanted to research into Mortgage Finance for the Ghanaian Worker. A research I wanted to conduct to enable me to develop a proposal on mortgage finance. I was very clear in my mind during discussions with Dr. Wisniewski that I could develop a proposal separately from my dissertation to achieve that purpose. We discussed my second option, a study of the Macroeconomics factors that affect Stock Market returns. This topic ignited a lot of interest in him. He helped in shaping up the topic and got some references on the topic for me. I spent some time at the library reading around the topic before meeting my dissertation supervisor Dr. C.B Tse. He approved the topic and probed further into the areas I intended to cover, he gave me valuable advice in the process. He was very helpful in subsequent discussions when I was developing the proposal. He helped in clarifying the research questions and objectives as well as some aspects of the methodology. Dr. Tse’s comments on my Dissertation Proposal A.G.C form became I major source of reference throughout the period of the dissertation. His guidance was clear as to what I must do. The research questions and objectives of the dissertation were well defined and fulfilled at the end of the process. Some few changes were made from what was anticipated at the proposal stage. At the proposal stage, I considered total returns comprising capital gains/losses and dividends yield as the best way of measuring stock market performance. Dividend yield data was not available for

61

most period of the study. I consider this a limitation, but market index must not necessarily have dividend yield. Stock Indexes are computed differently from one Stock Exchange to another. I was also not clear on whether I was looking at individual stock returns or the composite Index at the proposal stage. The outcomes compared quite well with the expected results. Inflation was expected to have a negative relationship with the Ghana All Share Index, but it turned out to have positive relationship this might have resulted from the use of nominal returns. Nominal returns and inflation are likely to be positively correlated as inflation can drive nominal earnings and where real interest rates are negative due to high inflation, investors will seek solace in the stock market. to seek higher returns. This was experience in Zimbabwe where galloping inflation drove the local stock market up in local currency terms. The research was well planned and executed to a large extent. Data collection and literature review was done on time. I realized at the stage of methodology review that the software I had for the analysis – SPSS was inadequate. The scope of analysis involved unit roots, co-integration and error correction techniques among others used in the realm of econometrics. I spent considerable amount of time looking for the appropriate software and I spent over three months to get it. Leaning to use the software was a big challenge. I overcame that by concentrating my effort on the modules I needed to analyse my data and interpret the results. Data availability was a major problem of this research. Data on Fiscal balance and GDP were available only annually. The analysis was done on monthly basis so the annual data was repeated for twelve (12) months. I should have looked for the software used for data projection to ensure the annual data was projected randomly for each month. The use of real rates instead of nominal rates contributed to the results I got. I should have used real market returns rate at least.

62

The learning process was tremendous. The dissertation was the second half of my studies. It involved consolidating all what I learnt for the assignments and examinations. I acquire more analytical skills than in the first stage of my studies. I also delved more into the area of Economics and Econometrics. I acquired vital skills that I have started employing in my work. In fact I can pass the test for a Macroeconomic Analyst. It was a fulfilling experience. It is good our Tutors are not allowed to help with the methodology and analysis of data. This helped to develop skills in finding answers to critical questions. I however want to recommend that they are allowed to propose software, give guidance on how to use and review the findings to ensure that the research results and interpretations are relevant and consistent. I hope my study has added to existing knowledge about the fact that there are lags in the effect of changes in macroeconomic factors on the Ghanaian Economy and the Stock Market. The transmission mechanism of policy need further study to identify why the lags and what can be done to correct it.

63

9.0

References

Books Alan C. Shapiro (2006), Multinational Financial Management, John Wiley & Sons, INC. Eighth Edition David F. Swenson (2000) Pioneering Portfolio Management The Free Press, New York

Gwartney J.D , Stroup R.L, Sobel R.S and Macpherson D.A (2003) Economics: Private and Public Choice, 10th Edition, South - Western

Richard A. Brealey, Stewart C Myers and Franklin Allen (2006) Corporate Finance, Mc Graw Hill Irwin, Eighth Edition Steve Lumby and Chris Jones (2003) Corporate Finance, Theory and Practice, Thomson, London, Seventh Edition.

64

Institute of Statistical, Social and Economic Research (ISSER), University of Ghana, Legon, The State of the Ghanaian Economy in 2008, Sundel Services, Accra University of Leicester (2006) Module 1 Edition 10, MN7022/D Foundations of Financial Analysis, Learning Resources: England. University of Leicester (2007) Module 7, MN7031D International Finance, Edition 10, Learning Resources: England.

Journals

Adam, Anokye and Tweneboah, (2008) Macroeconomic Factors and Stock Market Movement: Evidence from Ghana, MPRA Paper No. 11256, October2008 Anari, Ali, and James Kolari (2001) Stock Prices and Inflation Journal of Financial Research Vol XXIV NO. 4 Anthony Kyereboah –Coleman and Kwame F. Agyire – Tettey (2008) Impact of Macroeconomic indicators on stock market performance. The case of Ghana Stock Exchange. Journal of Risk Finance, Vol. 9. No. 4 pages 365-378. Bodie Zvi (1976) Common Stocks as a hedge against inflation Journal of Finance Vol. 31 (2) Chen N.F., Roll R. and Ross S, (1986) Economic Forces and Stock Markets, Journal of Business, pgs 383-403 Chin-Hong Pauh and T.K Jayaraman (2007) Macroeconomic Activities and Stock Prices in South Pacific Island Economy, Journal of Economics and

Management1(2), pgs 229-244

65

Choudhry, T. (2001) Inflation and rates of return on stocks: evidence from high inflation Countries, Journal of International Financial Markets, Institutional

Investors and Money, pgs 75-96 Cumhur Erdem, Cem Arslan and Meziyet Erdem (2005) Effects of

macroeconomic variables in Istanbul Stock Exchange Indexes. Applied Financial Economics pgs 987-994 Eugene Fama, (1991) Efficient Capital Markets: II, Journal of Finance, pgs 15751617 Fama and Schwartz (1977) Asset returns and inflation, Journal of Financial Economics, pgs 115-146

Fama, (1981) Stock returns, real activity, inflation and money, American Economic Review, pgs 269-282 Flannery, M.J. and Protapapadakis A.A (2002) Macroeconomic factors do influence aggregate stock returns, The Review of Financial Studies, 13(3) 75181. Geske, R and Roll, R (1983), Fiscal and Monetary linkage between Stock Returns and Inflation, Journal of Finance pgs 1-33 Hendry D. F (1986) Economics modeling with co integrated variables: An overview. Oxford Bulletin of Economics and Statistics 43(3) 201-212 Jay Ritter (2004) Economic growth and equity returns Pacific Basin Finance Journal 490 -503

66

John Y. Campbell and Robert J. Shiller (1988) Economic Forces and Stock Markets, The journal of Political Economy, Vol. 95, Issue 5 (October 1998), pgs 1062-1088 John Campbell, Robert Shiller (1987) Cointegration and Tests of Present Values. The Journal of Political Economy Vol. 95. pgs 1062-1088 Katarina Jeselius (2003) The cointegrated VAR Model Econometric Methodology and Macroeconomic Applications (Working Paper) Komain Jiranyakul, (2008) Empirical Assessment of the Present Value Model of Stock Prices using the Data from Thailand’s Stock Market , Nida Economic Review, Vol. 3 No. 1, pgs 24-36 Mansor Ibrahim and Wan. S. W Yusoff (2001) Macroeconomic Variables, Exchange rate and Stock price: A Malaysian Perspective, IIUN Journal of Economics and management 9. No, 2 141-163 Melina Dritsaki (2005) Linkage between Stock market and Macroeconomic Fundamentals: Case Study of Athens Stock Exchange. Journal of Financial Management and Analysis 18(1) pgs 35-47 Michael Adler and Bernard Dumas (1984), Exposure to Currency Risk: Definition and Measurement. Financial Management Summer 1984 pgs 50.. Michael 2006. Nagaratnam Sreedharan (2004) A Vector Error Correction Model (VECM) of Stock Market Returns Pardy, Robert, (1992) Institutional Reform in Emerging Securities Markets, Policy Research Working Paper, The World Bank WPS 907 67 Papaioannou (2006). Exchange Rate Risk Measurement and Management: Issues and Approaches for Firms, IMF Working Paper, November

Poon, S and Taylor, S.J (1991) Macro economic factors and UK Stock Markets, Journal of Business Finance and Accounting, pgs 619- 636 Robert Gay (2008) Effect of macroeconomic Variables on Stock Market Returns for Four Emerging Economies: Brazil, Russia, India and China, International Business and Economic Research Journal, Robert F. Engel and C.W.J. Granger (1987) Co-Integration and Error Correction: Representation, estimation and Testing, Econometrica Vol. 55 No. 2 pp 251-279 Serkan Y. Kandir, (2008). Macroeconomic Variables, Firm Characteristics and Stock Returns: Evidence from Turkey. International Research Journal of Finance and Economics Issue 16 Smith K and Sims A, (1993) Stock Market Performance and Macroeconomic variables, Applied Financial Economics, pgs 55-60

Ramir C. Maysami, Lee C.Howe and Mohamad A. Hamzah (2004), Relationship between Macroeconomic variables and Stock Market Indices: Co integration Evidence from Stock Exchange of Singapore’s All S Sector Indices. Pengurusan pg.47-77 Twerefou D.K and Michael K. Nimo, (2005) The Impact of Macroeconomic Risk on Asset Prices in Ghana, 1997 – 2002, African Development Bank 2005 Utku Utkulu, How to estimate long run relationships in economics: An overview of recent developments, Dokuz Eylul University Working paper Jurnal

68

Wan Mansor Mahmood

and Nazihah Dimmiah (2009) Stock Market

and

Macroeconomic Variables: Evidence from the six Asian – Pacific Countries. International Research Journals and Finance and Economics pgs. 154-163 Wongbagpo, P and Sharma, S.C (2002) Stock Market and Macroeconomic fundamental dynamic interactions: ASEAN – 5 Countries, Journal of Asian Economics , pgs 27-51

Anon, (n.d), “Economic variables and stock market returns: evidence from India”

10.0

Appendices

Appendix A: Unit Root Tests

Null Hypothesis: GSE has a unit root Exogenous: Constant Lag Length: 0 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level -2.704533 -3.460739 -2.874804 -2.573917 Prob.* 0.0749

69

*MacKinnon (1996) one-sided p-values.

Augmented Dickey-Fuller Test Equation Dependent Variable: D(GSE) Method: Least Squares Date: 11/12/09 Time: 23:41 Sample (adjusted): 1991M02 2008M12 Included observations: 215 after adjustments Variable GSE(-1) C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic) Coefficient -0.063786 3.093824 0.033200 0.028661 16.57213 58497.34 -907.7274 7.314497 0.007393 Std. Error 0.023585 1.529424 t-Statistic -2.704533 2.022869 Prob. 0.0074 0.0443 0.307023 16.81485 8.462580 8.493935 8.475249 1.941587

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter. Durbin-Watson stat

Null Hypothesis: GDP has a unit root Exogenous: Constant Lag Length: 12 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. 0.264663 -3.462574 -2.875608 -2.574346 Prob.* 0.9759

Augmented Dickey-Fuller Test Equation Dependent Variable: D(GDP) Method: Least Squares Date: 11/12/09 Time: 23:42 Sample (adjusted): 1992M02 2008M12 Included observations: 203 after adjustments Variable GDP(-1) D(GDP(-1)) D(GDP(-2)) D(GDP(-3)) D(GDP(-4)) D(GDP(-5)) D(GDP(-6)) D(GDP(-7)) D(GDP(-8)) Coefficient 0.004573 -0.006391 -0.006391 -0.006391 -0.006391 -0.006391 -0.006391 -0.006391 -0.006391 Std. Error 0.017277 0.057128 0.057128 0.057128 0.057128 0.057128 0.057128 0.057128 0.057128 t-Statistic 0.264663 -0.111867 -0.111867 -0.111867 -0.111867 -0.111867 -0.111867 -0.111867 -0.111867 Prob. 0.7916 0.9110 0.9110 0.9110 0.9110 0.9110 0.9110 0.9110 0.9110

70

D(GDP(-9)) D(GDP(-10)) D(GDP(-11)) D(GDP(-12)) C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic)

-0.006391 -0.006391 -0.006391 -0.565930 -0.003606 0.301070 0.252996 0.194816 7.173155 51.25580 6.262569 0.000000

0.057128 0.057128 0.057128 0.063158 0.083464

-0.111867 -0.111867 -0.111867 -8.960575 -0.043204

0.9110 0.9110 0.9110 0.0000 0.9656 0.016749 0.225405 -0.367052 -0.138555 -0.274612 2.015487

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter. Durbin-Watson stat

Null Hypothesis: FISCAL has a unit root Exogenous: Constant Lag Length: 0 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -1.402818 -3.460739 -2.874804 -2.573917 Prob.* 0.5805

Augmented Dickey-Fuller Test Equation Dependent Variable: D(FISCAL) Method: Least Squares Date: 11/12/09 Time: 23:43 Sample (adjusted): 1991M02 2008M12 Included observations: 215 after adjustments Variable FISCAL(-1) C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic) Coefficient -0.024668 -0.189035 0.009154 0.004503 1.023585 223.1655 -309.0789 1.967899 0.162127 Std. Error 0.017585 0.107440 t-Statistic -1.402818 -1.759449 Prob. 0.1621 0.0799 -0.074465 1.025897 2.893757 2.925112 2.906426 1.979723

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter. Durbin-Watson stat

Null Hypothesis: FISCAL has a unit root Exogenous: Constant Lag Length: 0 (Automatic based on SIC, MAXLAG=14)

71

t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -1.402818 -3.460739 -2.874804 -2.573917

Prob.* 0.5805

Augmented Dickey-Fuller Test Equation Dependent Variable: D(FISCAL) Method: Least Squares Date: 11/12/09 Time: 23:43 Sample (adjusted): 1991M02 2008M12 Included observations: 215 after adjustments Variable FISCAL(-1) C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic) Coefficient -0.024668 -0.189035 0.009154 0.004503 1.023585 223.1655 -309.0789 1.967899 0.162127 Std. Error 0.017585 0.107440 t-Statistic -1.402818 -1.759449 Prob. 0.1621 0.0799 -0.074465 1.025897 2.893757 2.925112 2.906426 1.979723

Null Hypothesis: INTEREST has a unit root Exogenous: Constant Lag Length: 1 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -1.619225 -3.460884 -2.874868 -2.573951 Prob.* 0.4710

Augmented Dickey-Fuller Test Equation Dependent Variable: D(INTEREST) Method: Least Squares Date: 11/12/09 Time: 23:45 Sample (adjusted): 1991M03 2008M12

72

Included observations: 214 after adjustments Variable INTEREST(-1) D(INTEREST(-1)) C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic) Coefficient -0.015563 0.327816 0.461285 0.114110 0.105713 1.686340 600.0297 -413.9702 13.58926 0.000003 Std. Error 0.009612 0.064929 0.295219 t-Statistic -1.619225 5.048848 1.562519 Prob. 0.1069 0.0000 0.1197 0.031168 1.783227 3.896918 3.944104 3.915985 2.033122

Null Hypothesis: INFLATION has a unit root Exogenous: Constant Lag Length: 1 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -2.281634 -3.460884 -2.874868 -2.573951 Prob.* 0.1789

Augmented Dickey-Fuller Test Equation Dependent Variable: D(INFLATION) Method: Least Squares Date: 11/12/09 Time: 23:46 Sample (adjusted): 1991M03 2008M12 Included observations: 214 after adjustments Variable INFLATION(-1) D(INFLATION(-1)) C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic) Coefficient -0.041270 0.166896 0.957790 0.044909 0.035856 4.390351 4067.063 -618.7358 4.960662 0.007847 Std. Error 0.018088 0.067898 0.529575 t-Statistic -2.281634 2.458056 1.808602 Prob. 0.0235 0.0148 0.0719 -0.044720 4.471243 5.810615 5.857801 5.829682 2.042697

Null Hypothesis: EXCHANGE has a unit root Exogenous: Constant Lag Length: 3 (Automatic based on SIC, MAXLAG=14)

73

t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. 0.045087 -3.461178 -2.874997 -2.574019

Prob.* 0.9607

Augmented Dickey-Fuller Test Equation Dependent Variable: D(EXCHANGE) Method: Least Squares Date: 11/12/09 Time: 23:47 Sample (adjusted): 1991M05 2008M12 Included observations: 212 after adjustments Variable EXCHANGE(-1) D(EXCHANGE(-1)) D(EXCHANGE(-2)) D(EXCHANGE(-3)) C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic) Coefficient 0.000636 -0.680179 -0.411622 -0.188433 0.011922 0.323581 0.310510 0.074885 1.160810 251.1767 24.75580 0.000000 Std. Error 0.014110 0.069524 0.078696 0.068597 0.008675 t-Statistic 0.045087 -9.783422 -5.230500 -2.746958 1.374317 Prob. 0.9641 0.0000 0.0000 0.0065 0.1708 0.005486 0.090184 -2.322422 -2.243257 -2.290425 2.042473

Null Hypothesis: D(GSE) has a unit root Exogenous: None Lag Length: 0 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -14.59452 -2.575813 -1.942317 -1.615712 Prob.* 0.0000

Augmented Dickey-Fuller Test Equation Dependent Variable: D(GSE,2) Method: Least Squares Date: 11/13/09 Time: 00:00 Sample (adjusted): 1991M03 2008M12

74

Included observations: 214 after adjustments Variable D(GSE(-1)) R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat Coefficient -1.000000 0.500000 0.500000 16.85709 60526.43 -907.6528 2.000000 Std. Error 0.068519 t-Statistic -14.59452 Prob. 0.0000 3.24E-17 23.83953 8.492082 8.507811 8.498438

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter.

Null Hypothesis: D(GDP) has a unit root Exogenous: None Lag Length: 11 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -8.031221 -2.576403 -1.942399 -1.615659 Prob.* 0.0000

Augmented Dickey-Fuller Test Equation Dependent Variable: D(GDP,2) Method: Least Squares Date: 11/13/09 Time: 00:03 Sample (adjusted): 1992M02 2008M12 Included observations: 203 after adjustments Variable D(GDP(-1)) D(GDP(-1),2) D(GDP(-2),2) D(GDP(-3),2) D(GDP(-4),2) D(GDP(-5),2) D(GDP(-6),2) D(GDP(-7),2) D(GDP(-8),2) D(GDP(-9),2) D(GDP(-10),2) D(GDP(-11),2) R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat Coefficient -1.562757 0.562757 0.562757 0.562757 0.562757 0.562757 0.562757 0.562757 0.562757 0.562757 0.562757 0.562757 0.679433 0.660971 0.194717 7.241718 50.29025 2.000000 Std. Error 0.194585 0.186483 0.178012 0.169118 0.159729 0.149753 0.139062 0.127479 0.114731 0.100378 0.083596 0.062456 t-Statistic -8.031221 3.017740 3.161339 3.327603 3.523201 3.757915 4.046802 4.414525 4.905004 5.606382 6.731904 9.010524 Prob. 0.0000 0.0029 0.0018 0.0011 0.0005 0.0002 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.006897 0.334415 -0.377244 -0.181389 -0.298009

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter.

75

Null Hypothesis: D(FISCAL) has a unit root Exogenous: None Lag Length: 0 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -14.59452 -2.575813 -1.942317 -1.615712 Prob.* 0.0000

Augmented Dickey-Fuller Test Equation Dependent Variable: D(FISCAL,2) Method: Least Squares Date: 11/13/09 Time: 00:06 Sample (adjusted): 1991M03 2008M12 Included observations: 214 after adjustments Variable D(FISCAL(-1)) R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat Coefficient -1.000000 0.500000 0.500000 1.031020 226.4195 -309.6891 2.000000 Std. Error 0.068519 t-Statistic -14.59452 Prob. 0.0000 -3.96E-18 1.458083 2.903636 2.919365 2.909992

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter.

Null Hypothesis: D(INTEREST) has a unit root Exogenous: None Lag Length: 0 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -10.45987 -2.575813 -1.942317 -1.615712 Prob.* 0.0000

Augmented Dickey-Fuller Test Equation Dependent Variable: D(INTEREST,2) Method: Least Squares Date: 11/13/09 Time: 00:07 Sample (adjusted): 1991M03 2008M12 Included observations: 214 after adjustments

76

Variable D(INTEREST(-1)) R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat

Coefficient -0.678698 0.339348 0.339348 1.688934 607.5820 -415.3085 2.025256

Std. Error 0.064886

t-Statistic -10.45987

Prob. 0.0000 0.000187 2.077908 3.890734 3.906463 3.897090

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter.

Null Hypothesis: D(INFLATION) has a unit root Exogenous: None Lag Length: 0 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level 10% level *MacKinnon (1996) one-sided p-values. -12.59545 -2.575813 -1.942317 -1.615712 Prob.* 0.0000

Augmented Dickey-Fuller Test Equation Dependent Variable: D(INFLATION,2) Method: Least Squares Date: 11/13/09 Time: 00:09 Sample (adjusted): 1991M03 2008M12 Included observations: 214 after adjustments Variable D(INFLATION(-1)) R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat Coefficient -0.853802 0.426873 0.426873 4.423428 4167.711 -621.3515 2.029338 Std. Error 0.067787 t-Statistic -12.59545 Prob. 0.0000 0.003224 5.842968 5.816369 5.832098 5.822725

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter.

Null Hypothesis: D(EXCHANGE) has a unit root Exogenous: None Lag Length: 2 (Automatic based on SIC, MAXLAG=14) t-Statistic Augmented Dickey-Fuller test statistic Test critical values: 1% level 5% level -12.87578 -2.575916 -1.942331 Prob.* 0.0000

77

10% level *MacKinnon (1996) one-sided p-values.

-1.615703

Augmented Dickey-Fuller Test Equation Dependent Variable: D(EXCHANGE,2) Method: Least Squares Date: 11/13/09 Time: 00:09 Sample (adjusted): 1991M05 2008M12 Included observations: 212 after adjustments Variable D(EXCHANGE(-1)) D(EXCHANGE(-1),2) D(EXCHANGE(-2),2) R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat Coefficient -2.208807 0.550403 0.167253 0.765762 0.763520 0.075508 1.191599 248.4018 2.027093 Std. Error 0.171547 0.128172 0.068310 t-Statistic -12.87578 4.294249 2.448428 Prob. 0.0000 0.0000 0.0152 0.000146 0.155273 -2.315111 -2.267612 -2.295913

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter.

**Appendix B Cointegration Tests
**

Date: 11/12/09 Time: 22:14 Sample (adjusted): 1992M05 2008M12 Included observations: 200 after adjustments Trend assumption: Linear deterministic trend Series: GSE GDP FISCAL INTEREST INFLATION EXCHANGE Exogenous series: GSE Warning: Critical values assume no exogenous series Lags interval (in first differences): 10 to 15 Unrestricted Cointegration Rank Test (Trace) Hypothesized No. of CE(s) None * At most 1 At most 2 At most 3 At most 4 Trace Statistic 7252.115 43.38458 22.92592 10.13615 4.425093 0.05 Critical Value 95.75366 69.81889 47.85613 29.79707 15.49471

Eigenvalue 1.000000 0.097235 0.061947 0.028151 0.021875

Prob.** 1.0000 0.8780 0.9622 0.9782 0.8664

78

At most 5

7.84E-06

0.001569

3.841466

0.9664

Trace test indicates 1 cointegrating eqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized No. of CE(s) None * At most 1 At most 2 At most 3 At most 4 At most 5 Max-Eigen Statistic 7208.731 20.45866 12.78977 5.711057 4.423524 0.001569 0.05 Critical Value 40.07757 33.87687 27.58434 21.13162 14.26460 3.841466

Eigenvalue 1.000000 0.097235 0.061947 0.028151 0.021875 7.84E-06

Prob.** 1.0000 0.7241 0.8962 0.9880 0.8122 0.9664

Max-eigenvalue test indicates 1 cointegrating eqn(s) at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I): GSE 0.086428 -0.015378 -0.007799 0.005749 -0.002715 0.004471 GDP -6.48E-17 1.369137 0.958362 0.081881 -0.432681 -1.113663 FISCAL 1.68E-17 0.094127 0.359203 -0.034294 0.019389 0.105148 INTEREST 1.82E-17 -0.025806 0.087470 -0.023885 -0.091327 -0.078724 INFLATION 1.92E-18 0.025669 -0.014088 -0.088472 0.007549 -0.022024 EXCHANGE 1.14E-15 -3.641976 2.115365 -1.769332 1.295989 -0.795081

Unrestricted Adjustment Coefficients (alpha): D(GSE) D(GDP) D(FISCAL) D(INTEREST) D(INFLATION) D(EXCHANGE) -11.57028 -0.030146 0.040699 -0.025498 -0.428142 -0.000442 4.86E-16 -0.022179 -0.029287 0.192802 -0.642663 0.004460 0.000000 0.010309 -0.059222 -0.214564 -0.486953 -0.008441 -6.25E-17 0.002382 -0.100283 -0.081938 0.051840 0.003766 -6.31E-16 0.019208 -0.056497 0.127323 0.105047 -0.001092 3.63E-16 -0.000204 -0.000494 0.001117 0.002208 -0.000114

1 Cointegrating Equation(s):

Log likelihood

2103.563

Normalized cointegrating coefficients (standard error in parentheses) GSE GDP FISCAL INTEREST 1.000000 -4.19E-16 3.12E-16 2.14E-16 (3.8E-08) (7.2E-09) (2.9E-09) Adjustment coefficients (standard error in parentheses) D(GSE) -1.000000 (1.6E-09) D(GDP) -0.002605 (0.00116) D(FISCAL) 0.003518 (0.00528)

INFLATION -3.81E-18 (1.8E-09)

EXCHANGE 1.21E-14 (9.0E-08)

79

D(INTEREST) D(INFLATION) D(EXCHANGE)

-0.002204 (0.01025) -0.037004 (0.02077) -3.82E-05 (0.00041)

2 Cointegrating Equation(s):

Log likelihood

2113.792

Normalized cointegrating coefficients (standard error in parentheses) GSE GDP FISCAL INTEREST 1.000000 0.000000 3.41E-16 2.07E-16 (6.6E-09) (2.4E-09) 0.000000 1.000000 0.068749 -0.018849 (0.06129) (0.02233) Adjustment coefficients (standard error in parentheses) D(GSE) -1.000000 -6.17E-16 (1.6E-09) (2.6E-08) D(GDP) -0.002264 -0.030366 (0.00117) (0.01827) D(FISCAL) 0.003968 -0.040098 (0.00535) (0.08350) D(INTEREST) -0.005169 0.263972 (0.01032) (0.16102) D(INFLATION) -0.027121 -0.879894 (0.02062) (0.32162) D(EXCHANGE) -0.000107 0.006106 (0.00041) (0.00642)

INFLATION 4.04E-18 (1.8E-09) 0.018748 (0.01652)

EXCHANGE 1.10E-14 (8.7E-08) -2.660052 (0.81049)

3 Cointegrating Equation(s):

Log likelihood

2120.187

Normalized cointegrating coefficients (standard error in parentheses) GSE GDP FISCAL INTEREST 1.000000 0.000000 0.000000 8.38E-17 (2.4E-09) 0.000000 1.000000 0.000000 -0.043584 (0.02859) 0.000000 0.000000 1.000000 0.359796 (0.15051) Adjustment coefficients (standard error in parentheses) D(GSE) -1.000000 -9.91E-16 (1.7E-09) (3.1E-08) D(GDP) -0.002345 -0.020486 (0.00117) (0.02226) D(FISCAL) 0.004430 -0.096854 (0.00536) (0.10163) D(INTEREST) -0.003495 0.058342 (0.01026) (0.19451) D(INFLATION) -0.023323 -1.346572 (0.02042) (0.38731) D(EXCHANGE) -4.09E-05 -0.001983 (0.00041) (0.00776)

INFLATION 4.13E-17 (1.8E-09) 0.026262 (0.02110) -0.109288 (0.11110)

EXCHANGE 5.54E-15 (8.0E-08) -3.753379 (0.95919) 15.90316 (5.04924)

-6.49E-16 (7.0E-09) 0.001615 (0.00495) -0.024029 (0.02258) -0.058924 (0.04322) -0.235407 (0.08606) -0.002612 (0.00172)

80

4 Cointegrating Equation(s):

Log likelihood

2123.043

Normalized cointegrating coefficients (standard error in parentheses) GSE GDP FISCAL INTEREST 1.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 1.000000

INFLATION -9.50E-16 (1.8E-09) 0.541869 (0.24056) -4.365727 (1.97620) 11.83014 (5.44791)

EXCHANGE -4.09E-15 (6.0E-08) 1.254711 (8.27790) -25.43963 (68.0037) 114.9062 (187.470)

Adjustment coefficients (standard error in parentheses) D(GSE) -1.000000 -2.85E-15 (1.7E-09) (3.1E-08) D(GDP) -0.002331 -0.020291 (0.00118) (0.02228) D(FISCAL) 0.003853 -0.105065 (0.00533) (0.10089) D(INTEREST) -0.003966 0.051633 (0.01026) (0.19444) D(INFLATION) -0.023025 -1.342327 (0.02046) (0.38771) D(EXCHANGE) -1.93E-05 -0.001675 (0.00041) (0.00775)

-5.12E-16 (7.0E-09) 0.001534 (0.00497) -0.020590 (0.02248) -0.056114 (0.04333) -0.237185 (0.08641) -0.002741 (0.00173)

-1.39E-16 (1.8E-09) 0.001417 (0.00126) -0.002029 (0.00568) -0.021786 (0.01096) -0.027247 (0.02184) -0.000943 (0.00044)

5 Cointegrating Equation(s):

Log likelihood

2125.255

Normalized cointegrating coefficients (standard error in parentheses) GSE GDP FISCAL INTEREST 1.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000

INFLATION 0.000000 0.000000 0.000000 0.000000 1.000000

EXCHANGE 4.57E-15 (5.6E-08) -3.684719 (0.84149) 14.35637 (5.61631) 7.067903 (13.7584) 9.115549 (14.6348)

Adjustment coefficients (standard error in parentheses) D(GSE) -1.000000 1.53E-15 (1.7E-09) (3.2E-08) D(GDP) -0.002383 -0.028602 (0.00117) (0.02287) D(FISCAL) 0.004007 -0.080620 (0.00531) (0.10392) D(INTEREST) -0.004312 -0.003457 (0.01023) (0.20008) D(INFLATION) -0.023310 -1.387779 (0.02046) (0.40021) D(EXCHANGE) -1.63E-05 -0.001203 (0.00041) (0.00801)

-2.17E-16 (7.0E-09) 0.001906 (0.00494) -0.021686 (0.02245) -0.053645 (0.04323) -0.235148 (0.08647) -0.002763 (0.00173)

-1.51E-16 (2.5E-09) -0.000337 (0.00174) 0.003131 (0.00789) -0.033414 (0.01520) -0.036841 (0.03039) -0.000844 (0.00061)

-1.76E-17 (1.8E-09) -0.000780 (0.00124) 0.008528 (0.00562) 0.016182 (0.01082) -0.013429 (0.02165) -0.000108 (0.00043)

81

**Appendix C Vector Error Correction
**

Vector Error Correction Estimates Date: 11/16/09 Time: 20:43 Sample (adjusted): 1992M05 2008M12 Included observations: 200 after adjustments Standard errors in ( ) & t-statistics in [ ] Cointegrating Eq: GSE(-1) GDP(-1) CointEq1 1.000000 -4.19E-16 (3.8E-08) [-1.1e-08] 3.12E-16 (7.2E-09) [ 4.3e-08] 2.14E-16 (2.9E-09) [ 7.5e-08] -3.81E-18 (1.8E-09) [-2.1e-09] 1.21E-14 (9.0E-08) [ 1.3e-07] -47.49835

FISCAL(-1)

INTEREST(-1)

INFLATION(-1)

EXCHANGE(-1)

C

Error Correction: CointEq1

D(GSE) -1.000000 (1.6E-09) [-6.2e+08] 7.19E-16 (2.8E-16) [ 2.53123]

D(GDP) -0.002605 (0.00116) [-2.23989] -0.000152 (0.00089) [-0.17087] -0.000968 (0.00094) [-1.02778] 0.000434

D(FISCAL) 0.003518 (0.00528) [ 0.66682] 0.000775 (0.00405) [ 0.19150] 0.002059 (0.00427) [ 0.48196] 0.024626

D(INTEREST) -0.002204 (0.01025) [-0.21502] -0.010068 (0.00786) [-1.28042] -0.011282 (0.00830) [-1.35936] -0.003405

D(INFLATION) -0.037004 (0.02077) [-1.78167] -0.005407 (0.01593) [-0.33937] -0.008492 (0.01682) [-0.50495] -0.024618

D(EXCHANGE) -3.82E-05 (0.00041) [-0.09390] 0.000310 (0.00031 ) [ 0.99365] 0.000116 (0.00033) [ 0.35350] 0.000166

D(GSE(-10))

D(GSE(-11))

6.90E-16 (3.0E-16) [ 2.30141] 4.65E-16

D(GSE(-12))

82

(3.0E-16) [ 1.54716] D(GSE(-13)) 3.65E-16 (3.0E-16) [ 1.22465] 6.11E-16 (3.0E-16) [ 2.04263] 5.09E-16 (3.0E-16) [ 1.70473] 2.14E-14 (1.9E-14) [ 1.11663] 2.05E-14 (2.0E-14) [ 1.02847] -1.71E-14 (2.5E-14) [-0.68440] -2.50E-16 (2.5E-14) [-0.01017] 1.33E-14 (2.4E-14) [ 0.54358] 1.87E-14 (2.5E-14) [ 0.75988] 7.78E-15 (4.3E-15) [ 1.81274] 6.46E-15 (4.4E-15) [ 1.47951] -1.30E-14 (6.2E-15) [-2.10347] -9.70E-16 (5.2E-15) [-0.18725] -4.28E-15 (5.2E-15)

(0.00094) [ 0.45893] 0.000285 (0.00094) [ 0.30458] -0.000323 (0.00094) [-0.34343] -0.000413 (0.00094) [-0.44049] -0.016497 (0.06024) [-0.27384] 0.001861 (0.06268) [ 0.02969] -0.480877 (0.07867) [-6.11290] 0.007041 (0.07718) [ 0.09123] -0.012254 (0.07674) [-0.15968] 0.023830 (0.07747) [ 0.30759] 0.000543 (0.01349) [ 0.04026] -0.001657 (0.01372) [-0.12080] 0.022213 (0.01936) [ 1.14742] -0.002467 (0.01629) [-0.15147] 0.008130 (0.01647)

(0.00428) [ 5.74768] -0.002128 (0.00424) [-0.50153] -0.000651 (0.00427) [-0.15245] -0.001746 (0.00425) [-0.41061] 0.078959 (0.27319) [ 0.28902] 0.016807 (0.28425) [ 0.05913] -0.800928 (0.35674) [-2.24510] 0.067641 (0.35000) [ 0.19326] -0.007006 (0.34802) [-0.02013] -0.055932 (0.35134) [-0.15920] -0.019736 (0.06120) [-0.32250] -0.026801 (0.06221) [-0.43082] -0.109272 (0.08779) [-1.24465] -0.028709 (0.07387) [-0.38862] -0.011928 (0.07467)

(0.00832) [-0.40904] 0.012899 (0.00824) [ 1.56457] 0.013997 (0.00829) [ 1.68813] 0.014149 (0.00826) [ 1.71242] -0.396103 (0.53080) [-0.74624] -0.355297 (0.55229) [-0.64332] -1.949335 (0.69313) [-2.81236] 0.110381 (0.68004) [ 0.16232] 0.280976 (0.67618) [ 0.41554] 0.502816 (0.68264) [ 0.73658] -0.003457 (0.11890) [-0.02907] -0.099805 (0.12087) [-0.82574] -0.073174 (0.17058) [-0.42898] -0.119251 (0.14353) [-0.83085] -0.030489 (0.14508)

(0.01687) [-1.45935] -0.007515 (0.01671) [-0.44981] -0.003076 (0.01680) [-0.18308] 0.001238 (0.01674) [ 0.07395] -0.871318 (1.07563) [-0.81005] -0.355861 (1.11917) [-0.31797] -8.696853 (1.40458) [-6.19176] -0.555260 (1.37805) [-0.40293] 0.066474 (1.37023) [ 0.04851] 1.292227 (1.38332) [ 0.93415] -0.038714 (0.24094) [-0.16068] -0.323720 (0.24493) [-1.32169] -1.074510 (0.34566) [-3.10855] -0.430108 (0.29085) [-1.47879] -0.291839 (0.29399)

(0.00033) [ 0.50333] 0.000630 (0.00033) [ 1.92561] -0.000992 (0.00033) [-3.01502] 0.000260 (0.00033) [ 0.79281] 0.005712 (0.02106) [ 0.27126] 0.008058 (0.02191) [ 0.36777] 0.002579 (0.02750) [ 0.09379] 0.064677 (0.02698) [ 2.39735] -0.069348 (0.02683) [-2.58515] -0.003448 (0.02708) [-0.12731] 0.000555 (0.00472) [ 0.11770] 0.002243 (0.00480) [ 0.46782] -0.003135 (0.00677) [-0.46322] 0.005912 (0.00569) [ 1.03827] -0.006676 (0.00576)

D(GSE(-14))

D(GSE(-15))

D(GDP(-10))

D(GDP(-11))

D(GDP(-12))

D(GDP(-13))

D(GDP(-14))

D(GDP(-15))

D(FISCAL(-10))

D(FISCAL(-11))

D(FISCAL(-12))

D(FISCAL(-13))

D(FISCAL(-14))

83

[-0.81770] D(FISCAL(-15)) -5.59E-15 (5.4E-15) [-1.03711] 2.04E-15 (2.9E-15) [ 0.70592] -1.67E-15 (3.0E-15) [-0.55825]

[ 0.49376] 0.008333 (0.01694) [ 0.49198] 0.005415 (0.00908) [ 0.59615] -0.004287 (0.00943) [-0.45465]

[-0.15974] -0.016999 (0.07681) [-0.22130] -0.002938 (0.04119) [-0.07133] -0.012522 (0.04276) [-0.29284]

[-0.21016] -0.048210 (0.14925) [-0.32302] -0.011812 (0.08003) [-0.14760] -0.002783 (0.08308) [-0.03349]

[-0.99269] -0.122737 (0.30244) [-0.40582] 0.151722 (0.16217) [ 0.93557] -0.026305 (0.16836) [-0.15625]

[-1.16001] -0.001022 (0.00592) [-0.17261] 0.000331 (0.00317) [ 0.10421] 0.002317 (0.00330) [ 0.70311]

D(INTEREST(-10))

D(INTEREST(-11))

D(INTEREST(-12))

3.70E-15 (3.0E-15) [ 1.23184] 4.65E-16 (3.0E-15) [ 0.15717] 1.29E-15 (3.0E-15) [ 0.43184] -2.21E-15 (3.0E-15) [-0.74689] -2.27E-15 (1.3E-15) [-1.81149] -1.34E-15 (1.4E-15) [-0.93810] 1.68E-15 (1.4E-15) [ 1.16317] 4.08E-16 (1.4E-15) [ 0.28122] -1.41E-15 (1.5E-15) [-0.96771] -2.47E-16 (1.5E-15) [-0.16949] 5.18E-14 (8.4E-14)

0.011480 (0.00943) [ 1.21719] -0.011573 (0.00929) [-1.24538] 0.002931 (0.00938) [ 0.31238] 0.005779 (0.00930) [ 0.62149] 0.001634 (0.00395) [ 0.41402] 0.008373 (0.00449) [ 1.86437] -0.023058 (0.00455) [-5.06860] 0.003059 (0.00456) [ 0.67126] 0.002205 (0.00459) [ 0.48071] 0.000674 (0.00458) [ 0.14712] 0.418113 (0.26484)

0.009621 (0.04277) [ 0.22493] -0.047040 (0.04214) [-1.11626] 0.013053 (0.04255) [ 0.30680] -0.039280 (0.04217) [-0.93153] -0.010695 (0.01790) [-0.59742] -0.019505 (0.02037) [-0.95770] 0.064343 (0.02063) [ 3.11889] -0.001759 (0.02067) [-0.08513] 0.002746 (0.02080) [ 0.13205] 0.012964 (0.02076) [ 0.62436] -0.471814 (1.20104)

0.100497 (0.08310) [ 1.20932] 0.016120 (0.08188) [ 0.19688] -0.038140 (0.08266) [-0.46138] 0.010323 (0.08193) [ 0.12600] 0.027787 (0.03478) [ 0.79889] 0.046215 (0.03957) [ 1.16794] -0.070728 (0.04008) [-1.76455] -0.067503 (0.04015) [-1.68114] -0.080652 (0.04041) [-1.99595] -0.067805 (0.04034) [-1.68078] 2.366684 (2.33355)

-0.148226 (0.16840) [-0.88021] -0.096941 (0.16592) [-0.58427] -0.196962 (0.16751) [-1.17581] 0.010184 (0.16602) [ 0.06134] 0.073150 (0.07048) [ 1.03783] 0.100027 (0.08019) [ 1.24745] -0.334116 (0.08123) [-4.11346] 0.040414 (0.08137) [ 0.49668] 0.005960 (0.08188) [ 0.07278] -0.061624 (0.08175) [-0.75382] 7.854994 (4.72878)

-0.002291 (0.00330) [-0.69504] 0.008141 (0.00325) [ 2.50625] -0.009235 (0.00328) [-2.81591] 0.002268 (0.00325) [ 0.69791] -0.001317 (0.00138) [-0.95463] -0.001202 (0.00157) [-0.76595] -0.000947 (0.00159) [-0.59553] 0.010353 (0.00159) [ 6.49926] -0.008943 (0.00160) [-5.57879] -0.000191 (0.00160) [-0.11950] -0.048635 (0.09258)

D(INTEREST(-13))

D(INTEREST(-14))

D(INTEREST(-15))

D(INFLATION(-10))

D(INFLATION(-11))

D(INFLATION(-12))

D(INFLATION(-13))

D(INFLATION(-14))

D(INFLATION(-15))

D(EXCHANGE(-10))

84

[ 0.61520] D(EXCHANGE(-11)) -1.80E-14 (1.0E-13) [-0.17806] -3.25E-14 (1.0E-13) [-0.31835] -7.44E-14 (9.5E-14) [-0.78129] -3.74E-14 (7.7E-14) [-0.48820] -2.66E-14 (6.0E-14) [-0.44652] -47.49835 (1.8E-14) [-2.6e+15] 1.000000 (3.7E-16) [ 2.7e+15] 1.000000 1.000000 5.90E-25 6.05E-14 4.34E+29

[ 1.57872] -0.249798 (0.31712) [-0.78771] -0.113871 (0.32095) [-0.35480] -0.029103 (0.29952) [-0.09717] 0.041360 (0.24089) [ 0.17170] 0.009984 (0.18735) [ 0.05329] -0.111118 (0.05678) [-1.95713] 0.002735 (0.00115) [ 2.37736] 0.431654 0.297510 5.832479 0.190333 3.217848 69.69982 -0.306998 0.336174 0.017000 0.227088 1.07E-29 2.92E-30 2038.733 -17.98733 -14.02935

[-0.39284] -0.602574 (1.43811) [-0.41900] -0.502796 (1.45546) [-0.34545] -0.188057 (1.35828) [-0.13845] 0.151280 (1.09242) [ 0.13848] 0.088028 (0.84964) [ 0.10361] 0.060797 (0.25748) [ 0.23613] -0.002456 (0.00522) [-0.47080] 0.355578 0.203478 119.9482 0.863145 2.337793 -232.6620 2.716620 3.359792 -0.048500 0.967131

[ 1.01420] 3.992407 (2.79416) [ 1.42884] 2.163403 (2.82788) [ 0.76503] -0.621776 (2.63906) [-0.23560] -2.006919 (2.12251) [-0.94554] -1.313086 (1.65079) [-0.79543] -0.028536 (0.50026) [-0.05704] -0.000518 (0.01014) [-0.05113] 0.272909 0.101298 452.8057 1.677038 1.590273 -365.5023 4.045023 4.688195 -0.003650 1.769030

[ 1.66110] -19.14410 (5.66219) [-3.38105] -13.91016 (5.73050) [-2.42739] -10.36970 (5.34788) [-1.93903] -6.894862 (4.30113) [-1.60303] -3.928232 (3.34521) [-1.17429] -2.053467 (1.01374) [-2.02563] 0.043512 (0.02054) [ 2.11823] 0.558877 0.454761 1859.414 3.398405 5.367835 -506.7576 5.457576 6.100748 -0.080350 4.602373

[-0.52534] -0.079326 (0.11085) [-0.71561] 0.424000 (0.11219) [ 3.77936] -0.029424 (0.10470) [-0.28104] -0.015817 (0.08420) [-0.18783] -0.005145 (0.06549) [-0.07856] 0.005762 (0.01985) [ 0.29032] -3.29E-05 (0.00040) [-0.08184] 0.584641 0.486606 0.712664 0.066532 5.963584 279.9186 -2.409186 -1.766014 0.005793 0.092855

D(EXCHANGE(-12))

D(EXCHANGE(-13))

D(EXCHANGE(-14))

D(EXCHANGE(-15))

C

GSE

R-squared Adj. R-squared Sum sq. resides S.E. equation F-statistic Log likelihood Akaike AIC Schwarz SC Mean dependent S.D. dependent

0.308450 17.43455

Determinant resid covariance (dof adj.) Determinant resid covariance Log likelihood Akaike information criterion Schwarz criterion

**Appendix D Granger Causality Tests
**

Pairwise Granger Causality Tests Date: 12/15/09 Time: 02:28 Sample: 1991M01 2008M12

85

Lags: 15 Null Hypothesis: D(GDP) does not Granger Cause D(GSE) D(GSE) does not Granger Cause D(GDP) D(FISCAL) does not Granger Cause D(GSE) D(GSE) does not Granger Cause D(FISCAL) D(INTEREST) does not Granger Cause D(GSE) D(GSE) does not Granger Cause D(INTEREST) D(INFLATION) does not Granger Cause D(GSE) D(GSE) does not Granger Cause D(INFLATION) D(EXCHANGE) does not Granger Cause D(GSE) D(GSE) does not Granger Cause D(EXCHANGE) D(FISCAL) does not Granger Cause D(GDP) D(GDP) does not Granger Cause D(FISCAL) D(INTEREST) does not Granger Cause D(GDP) D(GDP) does not Granger Cause D(INTEREST) D(INFLATION) does not Granger Cause D(GDP) D(GDP) does not Granger Cause D(INFLATION) D(EXCHANGE) does not Granger Cause D(GDP) D(GDP) does not Granger Cause D(EXCHANGE) D(INTEREST) does not Granger Cause D(FISCAL) D(FISCAL) does not Granger Cause D(INTEREST) D(INFLATION) does not Granger Cause D(FISCAL) D(FISCAL) does not Granger Cause D(INFLATION) D(EXCHANGE) does not Granger Cause D(FISCAL) D(FISCAL) does not Granger Cause D(EXCHANGE) D(INFLATION) does not Granger Cause D(INTEREST) D(INTEREST) does not Granger Cause D(INFLATION) D(EXCHANGE) does not Granger Cause D(INTEREST) D(INTEREST) does not Granger Cause D(EXCHANGE) D(EXCHANGE) does not Granger Cause D(INFLATION) D(INFLATION) does not Granger Cause D(EXCHANGE) Obs 200 F-Statistic 0.00737 0.17419 8.47020 4.45240 1.96324 1.03754 1.84887 0.81252 3.49212 2.31336 0.02838 0.00673 0.72316 2.50056 0.94117 1.54747 0.16560 4.25301 0.45075 0.75964 1.96425 4.56255 0.31808 2.37359 2.46927 0.19328 1.40261 1.90364 5.73895 11.8985 Prob. 1.0000 0.9998 3.E-14 5.E-07 0.0205 0.4194 0.0318 0.6627 3.E-05 0.0051 1.0000 1.0000 0.7588 0.0023 0.5200 0.0937 0.9999 1.E-06 0.9609 0.7204 0.0205 3.E-07 0.9931 0.0040 0.0027 0.9996 0.1509 0.0258 2.E-09 1.E-19

200

200

200

200

200

200

200

200

200

200

200

200

200

200

Appendix F: Proposal Template

86

**Electronic AGC FORM (PROPOSAL)
**

SECTION 1: STUDENT TO COMPLETE

NAME: I.D. No: ENROLMENT/START DATE

**SAMUEL YAO DAGADU
**

PROGRAMME:

069018828

DATE SUBMITTED 29/05/2009

04/07

LOCAL RESOURCE CENTRE QDL – GHANA

MSc FINANCE

MODULE: THE PROPOSAL

STUDENT DECLARATION: In submitting work to the University you are agreeing to the following statement: “I declare that this assignment is my own work, that all sources of reference are acknowledged in full and that it has not been submitted for any other course“.

**SECTION 2: TUTOR’S COMMENTS
**

Ability to construct a project with clear, coherent and well defended research questions/ objectives

Discussion of the relation between your proposed research and previous research

Discussion and justification of proposed methods

Overall Comments:

Second Marker Additional Comments (Optional):

87

**SECTION 3: TUTOR’S COMMENTS - Ethical Review Process:
**

Ethics Approval Decision Route: (Delete as appropriate)

1. None Required (student not doing research on live human subjects) 2. Automatic 3. Committee [School] [Faculty] [University] 1. Yes 2. No Comment:

Does this Project have Ethics Approval? (Delete as appropriate)

Tutor marking this assignment

Date of marking

Mark Awarded (%)

Grade Awarded

88

**School of Management Dissertation Proposal Pro-Forma Version 1.3 (November 2008)
**

Section 1: The Proposal Template

Your Name, Programme of Study, Student Number, Centre & Intake. SAMUEL YAO DAGADU, MSc FINANCE, 069018828, QDL-GHANA, APRIL 2007 Please identify any University of Leicester Tutors with whom you have discussed your proposal and the forum you used (e.g. workshops/Blackboard)

**Dr. C B TSE. SUMMER SCHOOL, 2008
**

Title (max. 15 words) Macroeconomic Variables and Stock Market Performance: Case study of Ghana Stock Exchange Abstract (max. 200 words)

This proposal is a prelude to a study of short run and long run relationships between macroeconomic variables and returns of listed stocks on the Ghana Stock Exchange using the Granger causality, Co integration techniques and Error Correction Models. Macroeconomic variables of interest are Fiscal deficit/surpluses, Inflation, Interest rates and Exchange rates. Economic factors that impact on changing investment opportunities and dividends directly impact pricing and performance of listed equity securities on the Ghana Stock exchange. The Government of Ghana has a key role to play in ensuring a sound capital market through macroeconomic measures, fiscal environment- taxation; legal, regulatory and institutional infrastructure. This study is important for the formulation of both fiscal and monetary policy. Individuals, Pension Funds and other Institutional Investors invest in the Stock market to ensure future stream of income. Erosion of capital in the market due to macroeconomic in-balances has dire consequences for both individual stock holders and Fund participants. Earlier studies on the Ghana Stock Exchange, reported on the impact of inflation, interest rates, foreign direct investments, exchange rates etc on the market. No study considered the effect of fiscal imbalances on the Ghana Stock Market.

Introduction (approx. 200 words)

89

The Ghanaian economy had experienced macroeconomic instability over long periods of time. Macroeconomic instability had negative effects on the Ghana Stock Exchange. It is not clear from available statistics and earlier studies on the Ghana Stock Exchange, the macroeconomic factor or combination of factors that are responsible for the performance of the exchange. In 1993 for instance, Inflation rose to 70% when, Interest rates was 30.95% and the Stock exchange recorded a performance rate of 116.06%. GDP grew in that year by 4.9%. In 1999, Inflation was as low as 21.3% and Interest rate was 34.19% but the stock performance was (15.14) % and GDP grew by 4.4%. The question then is: 1: What macroeconomic factors drive the Ghana Stock exchange? 2: How does fiscal deficits/surpluses, inflation rates, interest rates and exchange rates impact on the Ghana Stock Exchange? Finding answers to these questions will help add to existing knowledge about the underlying causes of price movement of the Ghana Stock Exchange and how these variables can be used to predict market returns. It will be useful in giving policy guidelines to ensure stability of the capital market. It will also be a useful guide to Investors and Financial Analyst in assessing the price of stocks and their systematic risks with anticipated changes in the macroeconomic factors.

Relation to previous research (approx. 400 words)

Fama (1981), Smith and Sims (1993), identified inflation, money supply, exchange rates as some of the major determinants of stock prices. Chen, Roll and Ross (1986) established the existence of long term equilibrium between stock prices and Inflation, Treasury–bill rate, Long term government bonds, Industrial production etc. Robert Pardy (1992), emphasised the role of macroeconomic and fiscal environment in the development and performance of securities market. According to Fama (1981) and other writers, inflation have negative relationship with stock performance because high inflation rates add to uncertainty which reduces business confidence and thus lowers stock returns, but studies on Ghana examined below indicate positive relationships with inflation. This is in line with Choudhry (2001) who investigated some developing Countries. Exchange rates have both positive and negative effect on stock prices depending on whether the stock is from a net import or export oriented company. Adam, Anokye and Tweneboah, (2008), examined the long and short term relationship between interest rate, inflation rate, net foreign direct investment, and exchange rate and the Ghana stock market during the period January 1991 to December 2006, using co integration and error correction models. They concluded that there is long run

90

relationship between stock prices and some of the macroeconomic factors examined. They found out that there is a positive relationship between inflation and share prices. Twerefou and Nimo (2005) investigated the impact of asset pricing of the various sectors of the Ghana Stock Exchange for the period January 1997 to December 2002 using arbitrage pricing method. The study concluded that inflation, short term interest rates and term structure of interest rates are macroeconomic factors affecting asset pricing in Ghana. Kyereboah-Coleman and Agyire-Tettey (2008) used quarterly time series data of the following macroeconomic factors: Inflation, Real Exchange rates, Interest rates, and Lending rates to examine the effect of these macroeconomic factors on the performance of the Ghana Stock Exchange. They concluded that lending rates have an adverse effect on stock performance and Inflation had a lagged effect on stock performance. They were certain on the fact that depreciation favours investors. This study will have a base on the macroeconomic relationships established with stock exchange performance by Fama, Chen Roll, Ross and others, but it will be based on the concepts of co integration and causality instead of Arbitrage Pricing Theory. Unlike all the studies on Ghana Stock Exchange, emphasis will not only be on what factors affect the stock market performance but also on causality of these factors. Moreover the effect of fiscal imbalance on the stock market will be examined.

91

Proposed methods (approx. 400 words)

Time series data from January 1991 to December 2008 will be used in the model. Data on Fiscal deficit and Inflation will be obtained from the Statistical Service of Ghana, Treasury bill rates and year on year exchange rate movements will be obtained from the Bank of Ghana. Ghana Stock Exchange All Share Index (GSI) which measures only price movements will be approximated by Databank Stock Index (DSI). The DSI is the depended variable of the regression equation. It represents the performance indicator of the stock market. Fiscal Deficits/Surpluses will capture the income and expenditure position of the economy; Interest rates will be approximated by 91 day Treasury bills; Inflation rate and Exchange rates are the other independent variables. The model is DSI = b0 + b1FD + b2INTrate + b3INFrate + b4EXrate + et

Eq1

where DSI is the Databank Stock Index, FD is Fiscal Deficit, INTrate is interest rate, INFrate is inflation rate, EXrate is the exchange rate, b0 - b4 are coefficients of variables and et is the error term, representing others factors not considered e.g. Oil price hike and change in Government. Visual inspection of the time series data will be done to observe and comment on movement in the variables. The time series data of each variable will undergo logarithmic transformation to stabilise the variances. Unit roots test using Augmented Dickey Fuller (ADF) will be undertaken to ascertain the order of integration of each variable to ensure stationarity of the variables. Granger causality tests using bivariate vector autoregressive method will be conducted to examine short term causal relationships between the stock returns and each variable by calculating F statistic. To establish long term relationships and co-movement among the variables, Co integration techniques by Granger (1986) and Engle and Granger (1987) will be used. The original time series data will be used to conduct an Ordinary least Squares regression between two variables at a time under the model Yt = a + Bxt + ut

Eq2

Where, ut is the estimated residuals on the long run equilibrium. The residuals are then subjected to unit root test. If co integration is confirmed with any variable, then Johansen and Juselius (1990) Error Correction Model will be used to establish long run causal relationship between the two variables. Multiple regression analysis alone will not be able to answer the research questions since the variables of interest are interdependent and the analysis will require a test of the stationarity of the data and establish the causality of price movement and dependencies of the variables. Arbitrage Pricing Theory will not be useful for this study.

Reflections (approx. 500 words)

92

The success of this work depends largely on the availability, and frequency of data; my ability to interpret the results and understand the real drivers of returns of the Ghana Stock Exchange. Total returns comprising capital gains/losses and dividends yield is the best way of measuring stock market performance but dividend yield data may not be available for most period of the study. This will limit stock returns to only capital gains or price changes. This may not show the true performance of the Stock Market, because it is possible for some companies to post good financial performance and pay high dividends during periods of low patronage of the Stock exchange, but because of lack of investor interest in the stock market, prices of these companies will not reflect the economic fundamentals but rather follow the general trend of a bearish market. Timeliness and frequency of data, in a study of this nature is paramount. It will be ideal to use monthly time series data but information on Fiscal deficits for instance, which is key in this analysis cannot be obtained monthly. Using annual data will give us only eighteen (18) regression points (1991-2008) as against two hundred and sixteen (216) points using monthly data. This may limit the fine details required for effective time series analysis. The analysis may go beyond the capability of Excel, it will involve finding logs, unit roots, cointegration and error correction techniques used in the realm of econometrics. My ability to interpret the results will depend on how quickly I grasp the techniques involved. The Economy of Ghana like most developing countries defy the principles of economics espoused in text books. Inflation can for instance be more than the rates for treasury instruments and yet treasury instruments will be considered more viable than investment in the capital markets. I hope this study will confirm the results of other studies on macroeconomic relationship with stock markets in developing countries. Government’s fiscal and monetary policies have significant effect on the economy and therefore the Capital Market. Other factors like oil price hikes; change of Government; international financial and economic development, have major impact on the economy of Ghana, but these do not form part of this study. It is difficult to say whether the Ghana Stock Exchange is efficient or not, it is difficult to depend on past price movements to make gains but the same cannot be said of release of information. Some Companies announce good returns and prospects but nothing happens to their prices.

Government’s debt financing cause interest rates to rise and this directly cause Inflation and depreciation. Depreciation results in capital flights. Investors not having confidence in the economy divert resources from the long term investments to short term treasury bills and consumables or real estate. Listed Companies are staved of needed resources to finance viable projects since they cannot borrow at the prevailing high interest rates. They become less competitive and their profit levels fall leading to a fall in returns.

Conclusion (max. 200 words)

Government have a key role in creating macroeconomic disequilibrium leading to high inflation and interest rates, devaluations and fiscal deficits among others which impact negatively on the stock market. 93

Participants in the stock market anticipate real returns from the stock market so stock prices will move inversely with inflation. Investors directly compare earning yield on stocks with treasury yields and move funds from one market to the other, an inverse relationship between stock returns and interest rates is expected. Fiscal deficits increase Government borrowing which increases interest rates, which crowd out private individuals and businesses, denying listed companies of much needed capital resulting in low returns. Depreciation of the cedi leads to decrease in the prices of exports and increase in demand, which leads to increase cash-flows. A weak currency increase the price of imported goods so Companies of a net importing Country like Ghana will suffer with currency depreciation. After submission of this proposal, I shall wait for the results to start data collection and writing of the dissertation. I shall in the interim continue with literature review, studying research methodology, and learn about the econometric analytical tools I shall employ to analyse the data.

Timetable (approx. 100 words, or a one page diagram)

AUG 2008

SEPT 08 TO APRIL 09

MAY 09

JUNE 09

JULY 09 WK 4

Event/ Date Agree topic with Supervisor Initial Literature Review Proposal Writing Proposal Submission (May 29) Literature Review Research Methodology Data Collection Learning of Analysis Methods Proceed on Leave Data Analysis Writing of Dissertation Results & Analysis Conclusions References Finalization Final Editing Proof-Reading Submission

Wk 3

SEP TO DEC 08

JAN TO APRIL

MAY

JUNE

WK1

WK2

WK3

WK5

94

References Journals

Adam, Anokye and Tweneboah, (2008) Macroeconomic Factors and Stock Market Movement: Evidence from Ghana, MPRA Paper No. 11256, October2008 Anthony Kyereboah –Coleman and Kwame F. Agyire – Tettey (2008) Impact of Macroeconomic indicators on stock market performance. The case of Ghana Stock Exchange. Journal of Risk Finance, Vol. 9. No. 4 pages 365-378. Chen N.F., Roll R. and Ross S, (1986) Economic Forces and Stock Markets, Journal of Business, pgs 383-403 Choudhry, T. (2001) Inflation and rates of return on stocks: evidence from high inflation Countries, Journal of International Financial Markets, Institutional Investors and Money, pgs 75-96 Fama and Schwert (1977) Asset returns and inflation, Journal of Financial Economics, pgs 115-146 Fama, (1981) Stock returns, real activity, inflation and money, American Economic Review, pgs 269-282 Fama, (1991) Efficient Capital Markets: II, Journal of Finance, pgs 1575-1618 Pardy, Robert, (1992) Institutional Reform in Emerging Securities Markets, Policy Research Working Paper, The World Bank WPS 907 Smith K and Sims A, (1993) Stock Market Performance and Macroeconomic variables, Applied Financial Economics, pgs 55-60 Twerefou D.K and Michael K. Nimo, (2005) The Impact of Macroeconomic Risk on Asset Prices in Ghana, 1997 – 2002, African Development Bank 2005 Utku Utkulu How to estimate long run relationships in economics: An overview of recent developments, Dokuz Eylul University Working paper

Appendices (optional)

95

Section 2: University of Leicester School of Management - Initial Ethical Review Form

**Leicester University School of Management Ethical Review Form: Part 1
**

Student Statement. I have read the above information. I confirm that my research does not involve the study of live human beings. I have read the above information. I confirm that my research does involve the study of live human beings. Insert X Student Action.

You do not need to complete Part 2 of this form. Ethics approval is not required.

Statement 1

X

Statement 2

Please proceed to complete Part 2 of this form.

You are only required to fill in part 2 of this form if your research involves studying live human beings. In cases of automatic ethics approval or where no ethics approval is necessary please allow 8-10 weeks from receipt by the University for the return of your grade. In instances where part 3 of the Ethics Form is completed you should allow 8-14 weeks. Proposals that are received without the completed Ethical Review Form will be returned to the student unmarked.

96

**Leicester University School of Management Ethical Review Form: Part 2
**

Please answer all of these questions by ticking yes or no in the box provided Yes 1. 2. 3. 4. 5. 6. Is physical pain or psychological stress from the proposed project likely to cause harm or negative consequences beyond the risks in normal life? Will the study involve prolonged or repetitive testing? 8. 9. 10. Will financial inducements (other than expenses) be offered to participants? Will the study involve recruitment of patients or staff through the NHS? Does the study involve participants who are particularly vulnerable or unable to give informed consent? (e.g. people under the age of 18, people with learning disabilities, students you teach or assess) Will it be necessary for participants to take part in the study without their knowledge and consent at the time? Does the study involve audio or visual recording of people in public places? Will the study involve the discussion of sensitive topics? (e.g. sexual activity, drug use, illegal activities, death, whistle blowing) Are drugs, placebos or other substances to be given to the study participants or will the study involve invasive, intrusive or potentially harmful procedures of any kind? Will blood or tissue samples be obtained from participants? No

X X X X X X X X X X

7.

If your answer is yes to any of these questions, please fill in Part 3.

97

**Leicester University School of Management Ethical Review Form: Part 3
**

In no more than a page – 1. Explain why you ticked yes to one or more of the questions on Part 2, and how you plan to address the ethical issues raised. You will need to do this in consultation with a Dissertation Tutor on Blackboard. Please identify which Tutor you discussed these issues with. Blackboard Tutor’s Name: -------------------------------------------------------------------------------------------------------

Assessor’s Comments (to be completed by the markers of the proposal)

Assessor’s Name: Assessor’s Signature: Date:

98

---------------------------------------------------------------------------------------------------------END OF STUDENT SUBMISSION THE FORMS ON THE FOLLOWING PAGES ARE ONLY USED IN INCIDENTS OF PLAGIARISM AND/OR THE AWARD OF A FAIL GRADE FOR THIS PIECE OF WORK. STUDENTS SHOULD NOT DELETE THESE FORMS.

99

School of Management

PLAGIARISM FORM (Student Assessed Work)

This form should only be completed where a case of plagiarism (inc. ‘poor scholarship’) has been identified. It should be attached to the student’s submitted work that contains the plagiarised material.

**PART 1 – TO BE COMPLETED BY FIRST MARKER
**

Student Name: Programme of Study & Year: Module: Nature and Extent of Plagiarism (please highlight or underline plagiarised text): What, if known, is the source of plagiarism? (e.g. peer plagiarism, single or

multiple texts, websites)

What proportion of the submission is plagiarised? Any other comments? What mark should be awarded? Name of 1st marker: Date (dd/mm/yyyy)

% %

**PART 2 – TO BE COMPLETED BY SECOND MARKER
**

Comments

(please indicate agreement or disagreement with the first markers assessment and recommended action):

Name of 2nd Marker Date (dd/mm/yyyy)

**PART 3 – TO BE COMPLETED BY DEPUTY HEAD OF SCHOOL
**

Recommendation

(following, where appropriate, a student interview and consultation with the Director of Postgraduate Studies and the Director of Undergraduate Studies)

Resubmission Permitted? Mark Awarded Name of Deputy Head of School Date (dd/mm/yyyy)

Yes/No %

100

School of Management

Second Marker Feedback Form

Student Name: Programme Date of Enrolment Module Is resubmission Yes/No2 automatically permitted? Your assignment has been subject to a system of second marking. Unfortunately the markers have concluded that your assignment has not met a satisfactory standard. You should follow the advice offered on the AGC form and consider the information below. REASON(S) FOR THE FAIL GRADE ACTION TO TAKE You need to check carefully that you have understood the You have not adequately answered the question set. Please discuss your interpretation of the question set. question with a Tutor on Blackboard and use the support materials on writing assignments found on Blackboard. You have not adequately explained Please discuss the expectations of the assignment with the what you have done. Tutor on Blackboard. Your answer is too descriptive – it lacks Please refer to the support material on Blackboard. Please sufficient analysis to address the discuss your plans for the resubmission with a Tutor on question set. Blackboard. Please keep to the word count stipulated in the assignment Your answer is too brief or exceeds the question. For advice on developing an essay please see your word limit set for this assignment. Programme Handbook and Blackboard. You do not make sufficient use of the Please discuss any plans for a resubmission with a Tutor on concepts and theories that are relevant Blackboard. to addressing the assignment question. You have made use of literature/study Please read the guidance in your programme handbook on materials without fully acknowledging referencing. There is further information on how to avoid the sources.* plagiarism on Blackboard. You have simply reproduced the information contained in the module Please read the guidance in your programme handbook on and other readings. You must use this referencing. There is further information on how to avoid material to answer the question in your plagiarism on Blackboard. own words.* Please read the guidance in your programme handbook on Your answer is too similar to that of referencing. There is further information on how to avoid another student.* plagiarism on Blackboard. Other: Please specify: Items marked with an * are serious academic offences and amount to plagiarism or cheating. Please see your Programme handbook about the regulations governing plagiarism. Second Marker: Date:

2

In cases where there is not automatic resubmission the case is presented at the Progress Board.

101

102

by samdagdivine

This study examines the long and short run relationships between macroeconomic variables and stock market returns in general using Ghana Stock Exchange Index as a special case, to find answers to t...

This study examines the long and short run relationships between macroeconomic variables and stock market returns in general using Ghana Stock Exchange Index as a special case, to find answers to the following questions: What macroeconomic factors drive the performance of Ghana Stock Index? How does GDP, Fiscal balance, Inflation, Interest rates and Exchange rates impact the Ghana Stock Index (GSI)? The macroeconomic variables used for this study are Gross Domestic Product (GDP), Fiscal balance (deficit/surpluses), Inflation, Interest rates and Exchange rates.

- Impact of Macro Economic Variables on Stock Prices
- Impact of Macro Economic Factor on Stock Market
- Macro Economic Determinants of Malaysian Stock Market
- 0011 Relationship Between Stock Market and Macro Economic Variables in India
- The Relationship Between Stock Market Volatility and Macro Economic Volatility
- Supply Chain Management of Wal-Mart
- Efficiency Hypothesis of the Stock Exchange Project Presented by MR CAMARA MOHAMED BINTOU
- Nishat Textile
- Impact of Macroeconomic Factors On Money Supply
- Micro Finance Thesis Report
- 552-1401-1-SM
- 501-513-libre.pdf
- Macroeconomic
- BBEK4203
- Macroeconomic Factors and Performance of Indian Stock Market
- The Review of Stock Returns and Macroeconomic Variables
- Macro Eco
- Zimbabwe
- ch01
- chapter 1
- ECON1020-Sem 1 2011 Final Exam
- Self Study
- The Macroeconomic Situation
- Long-run Relationship Between Islamic Stock Returns and Macroeonomic Variables.مهمة جدا في تفسيرك لمعنى العلاقة بين المتغيرات ص4
- Chapter 5 - Introduction to Macroeconomics WMN
- Chapter 1 Introduction to Macroeconomics
- macroeconomic analysis in malaysia.docx
- june 2007
- Project
- 2011 Nobel Prize in Economics

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd