i
DEDICATIONS
This dissertation is dedicated to my late father, and my mother for their enduring moral and
financial support.
Thank you, uncle, Dorcas and Ruvimbo
ii
ACKNOWLEDGEMENTS
I wish to ac
knowledge my profound gratitude to the many people
who contributed in various
ways to make this project a reality. The preparation of this document would not have been
possible without the invaluable and unequal support received from these people in their different
capacities.
Most of all I would like to
thank
the
L
ord for the guidance. In addition I would like to thank
my
supervisor,
Mr. Mandishekwa
for h
is
tremendous efforts and guidance that really helped me in
completion of this project. I would also like to thank my econometrics lecturer,
Mr.
Dzingira
i
who made me know this module
, that I can do a project on it. I would like to thank the entire
Economics department as a whole for guidance.
I would also want to acknowledge the tremendous support received from my family and friends.
To Sylvia Musekiwa a
nd Sikelela Ncube thank so mu
ch for your machine

you saved
iii
ABSTRACT
This study examines the dynamic interactions between the stock market and inflation, 90

day
Treasury
bill rate and exchange rate in Zimbabwe for the period 1995:01 to 2007:03.
The
study
uses tests for Granger causality. The stock market has been on an upward trend fo
r
most of the
period under study.
The depreciation of the currency on the parallel market and the changes in
inflation can explain the recent rapid increase in stock pr
ices.
Granger causality tests using the
Sims approach reveal uni

directional causality between inflation and industrial share index and
bi

directional between exchange rate and the industrial share index and bi

directional causality
from industrial share i
ndex to interest rates
iv
TABLE OF CONTENTS
DEDICATIONS
................................
................................
................................
.............................
i
ACKNOWLEDGEMENTS
................................
................................
................................
.........
ii
ABSTRACT
................................
................................
................................
................................
..
iii
TABLE OF CONTENTS
................................
................................
................................
............
iv
1.0
Introduction
................................
................................
................................
......................
1
1.3 Objectives of the Study
................................
................................
................................
.........
4
1.4 Significance of the Study
................................
................................
................................
......
5
1.5 Research Hypothesis
................................
................................
................................
.............
5
1.6 Research Questions
................................
................................
................................
...............
5
1.7 Assumptions
................................
................................
................................
..........................
6
1.9 Organization of the Study
................................
................................
................................
.....
6
CHAPTER 2
................................
................................
................................
................................
..
7
LITERATURE REVIEW
................................
................................
................................
............
7
2.0 Introduction
................................
................................
................................
...........................
7
2.1 Theoretical Review
................................
................................
................................
...............
7
2.2 Empirical Review
................................
................................
................................
..................
9
2.3 Conclusion
................................
................................
................................
...........................
14
METHODOLOGY
................................
................................
................................
.....................
15
3.0Introduction
................................
................................
................................
..........................
15
3.1 Model Specification
................................
................................
................................
............
15
3.2
Diagnostic Tests
................................
................................
................................
..................
17
3.2.1 Unit Roots Tests
................................
................................
................................
...............
17
3.2.2 Integration and Cointegration Tests
................................
................................
.................
17
3.2.3 Causality Tests
................................
................................
................................
.................
18
3.3 Justification of Variables
................................
................................
................................
.....
18
3.3.1 Exchange Rates
................................
................................
................................
.............
18
3.3.2 Consumer Price Index (As a Measure of Inflation)
................................
......................
18
3.3.3 Interest Rates
................................
................................
................................
................
19
v
3.3.4 The Stochastic Error Term (The Disturbance Term)
................................
....................
19
3.4 Data Characteristics
................................
................................
................................
.............
20
3.5 Strengths and Weaknesses of the Model
................................
................................
.............
20
3.5 Conclusion
................................
................................
................................
...........................
20
CHAPTER FOUR
................................
................................
................................
.......................
21
DATA PRESENTATION
................................
................................
................................
...........
21
4.0 Introd
uction
................................
................................
................................
.........................
21
4.1 Unit Roots Tests and Cointegration Results
................................
................................
.......
21
4.2 Interpretation of Results
................................
................................
................................
......
24
4.3 Conclusion
................................
................................
................................
...........................
25
CHAPTER FIVE
................................
................................
................................
........................
25
RECOMMENDATIONS and CONCLUSIONS
................................
................................
......
25
5.0 Introduction
................................
................................
................................
.........................
25
5.1 Policy Recommendations
................................
................................
................................
....
26
5.
2 Suggestions for Future Research
................................
................................
.........................
27
5.3 Conclusions
................................
................................
................................
.........................
28
REFERENCES
................................
................................
................................
............................
28
APPENDIX
................................
................................
................................
................................
..
33
vi
List of
T
ables
Table one

optimal lag
length……………………………………………………
..................21
Table two

results of unit roots tests……………………………………………..
.................22
Table 3

causality results…………………………………………………………..................23
Table four

direction of causality..............................
................................................................24
vii
List
Of Figures
Figure one:
T
rend in the industrial index…………………………………
2
Figure two:
T
rend of the inflationary pressure……………………………
3
viii
List of
appendi
x
Appendix A
:
data set
……………………………………………………
34
Appendix B unit roots tests
……………………………………………..
36
Appendix C causality tests
………………………………………………
40
1
CHAPTER ONE
INTRODUCTION
1.0
In
troduction
Since
stock marke
ts
are vital for
any well
defined financial system of a
country,
this
research
will
explore the linkages and the relationships between the Zimbabwe
S
tock
Exchange (
ZSE)
industrial
index and
the highlighted macroeconomic variables of
consumer
price
index (
CPI)
,
short
term nominal interest rates
(90

day TB

rate)
and exchange
rates
(ZWD/USD)
. The
Zimbabwe
stock exchange acts as a secondary market giving a platform for companies
to
raise
capital
t
hrough
‘
issues
’
to the
investing
public. This
chapter encompass
es: the background of the
study
,
statement of the problem
,
objectives
,
significance o
f the study, research questions
,
research
hypothesis, limitation
/
delimitations, assumptions
and
organization
of the study
.
1.1
Background
of
the
Study
The
Zimbabwe stock exchange is a small dynamic stock
exchange. It
was open to investors since
1993.
It
formalized its operations following passing of the
S
tock
E
xchange
A
ct which was
implemented in 1974 by then it was the Rhodesian
S
tock
E
xchange (
RSE).
S
ince then the
Zimbabwe
stock
exchange has grown immensely to become one of the most important equities
exchange in
Africa
and a provider of
services
that
ease the
increasing
of capital and the dealings
of
shares. In
Africa the Zimbabwe
stock exchange boast as the third largest bourse
after
the
Johannesburg
S
tock
E
xchange
(JSE) of
S
outh
Africa
and
Casablanca
S
tock
Exchange
of
Morocco. According
to its
website (
www.zs
e.co.zw
), more
than seventy five (75) local and
international companies are now listed with the local bourse.
The
Zimbabwe
stock exchange reports on trading prices daily .
B
asically the
re
are two weighted
ind
i
ces;
the industrial index and the mining index .
T
he industrial index uses data from various
industrial counters listed with the Zimbabwe stock exchange and it cuts across all the sectors of
the
economy. The
mining index on the other hand
uses
strictly data from the mining sector
of the
e
conomy (ZSE webs
ite).
2
According to Adept solutions (1999), developments in the Zimbabwe Stock Exchange have been
largely influenced by the prevailing unstable macroeconomic conditions, characterized by low
interest rates, dual interest rates, exchange control, shortage
of foreign currency and
hyperinflation as well as the unstable political situation, land reforms and decline in foreign
participation.
Over
the
years 199
5
to 200
5
the trend between the
Zimbabwe
S
tock exchange and the selected
macroeconomic variables of
in
terest
rates, exchange
rates and
consumer
price index was rather
positive
.
T
he industrial index was fluctuating below the inflatio
n rate (represented by the
CPI in
this case
)
and the interest rates from the beginning of the period under study but later cau
ght up
with inflation rate as from 2006.
B
esides the soaring interest rates and ever increasing inflation
rate as well as
persistent
depreciation in the
Zimbabwean
do
llar against major
currencie
s such as
the united states
dollar
, the
ZSE
sustained an upward
trend (
2006
monetary
policy statement).
The trend in the ZSE index is shown in the figure
below.
Figure One: Trend In Industrial Index
Source
Z
imbabwe Stock Exchange
It is due to the above
mentioned background
that The African Stock Exchange
Association
(
ASEA) rated the
ZSE
as the
best
performing
bourse on the African continent for the year 2005.
.
This
rise was well above
of the Casablanca
S
tock
E
xchange one of
its
closest rivals
.
One of the major challenges to the economy at the inception of reforms was high inflation. This
high level of inflation resulted in reduced operations and company closures because of cash flow
0
20
40
60
80
100
120
140
160
180
200
Index%
3
mismatches. The inflation level for the year 200
7
was very hig
h
compared to
SADC partners
thus eroding the economy’s export competitiveness. It was fuelled by high budget deficit, high
interest rates, deteriorating terms of trade, ele
ctricity and fuel shortages. (
RBZ
publications)
The
graph below illustrates the tren
d of the inflationary pressures that affected the economy.
Figure two
:
Trend Of The Inflationary Pressure
Source
: Reserve Bank Of Zimbabwe
The trend of the consumer price index indicates that there was an upward path of the inflation
between 199
5
and
20
04
. This upward trend in inflation continued even up to 2007.
These
inflationary trends are expected to have depressed stock market activity by lowering business
activity. Because it is the main signal for business in market driven economies, an increase in
inflation raises the country’s economic risk and thus can s
care off foreign investors.
When reforms were introduced the local currency had to be devalued to promote exports
.
The
local currency was then pegged at about Z$38/US$ from 1998 until it was devalued again by
44.4% in August 2000 to Z$55/US$
(RBZ 1997)
.
The influence that the devaluation of the
Zimbabwean dollar could have had on the performance of the ZSE is not clear. It depends on the
net benefits to both exporters and importers.
0
200
400
600
800
1000
1200
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
cpi
4
Interest rates were relatively high because of limited competition in th
e banking system. In
October 1995, the RBZ redefined the rediscount rate to reflect its medium term view on inflation
and thus maintained it at 29.5% when inflation was above
22 %
.
( RBZ
1996)
.
High inflation
resulted in the minimum bank lending rates being
raised negatively impacting on equity
investment. Positive real rates of interest were maintained in the money market
.
The impact
the
Treasury
bill rate could have had on the stock market is obviously negative
because an increase in the
Treasury
bill rate is expected to have to channel funds from the stock
market to the money market.
1.
2
Statement
of
the
Problem
Basically
according to theoretical and economic
postulations, a
goo
d stock market
performance
is
due to good macroeconomic
performance, Smith
et al
(1994
)
.
However
in the
case of
Zimbabwe, the
country was experiencing the worst melt down during most of the period under
review but the stock market has sustained an upw
ard
trend. According
to the
ZSE
website, it
was
rated as the as the best
performing
stock exchange in the
world
rising capital markets
in 2007
.
The researcher therefore intends to investigate why the
ZSE
under those unfor
tu
nate
circumstances has managed to defy the odds and emerge as the best
perfo
rming
stock exchange
among the world emerging capital markets by looking at the relationship between the
ZSE
’s
industrial index and
the
selected macroeconomic variables.
1.
3
O
bjectives
of
the
Study
In carrying out this
research, the
researcher intends to fulfill the following objectives:
To establish whether the selected macroeconomic variables are crucial in explaining the
behavior of the stock market.
Scrutinize
the long run relationship between the performance of
ZSE
and the selec
ted
macroeconomic variables(interest rates, exchange rate and consumer price index)
To contribute to the international debate regarding the relationship between these
variables.
In
so doing the researcher will avail some new
literature
on the sophisticated
relationship
among the selected macroeconomic variables and the stock market
industrial index
.
5
1.
4
S
ignificance
of
the
Study
In tackling this
topic, the
researcher will get an informed perspective between the
ZSE
industrial
index,
interest
rates,
exchange rates and consumer price
index. The
institution, Midlands
S
tate
U
niversity is going to harness this work as one of the foundations for further research to the
related areas as well as teaching
aids. The
information from this research can also be
lead to
effective policy implementation by policy
makers. To
the investor it aids investment decision in
different ec
onomic climate.
Although research in the area has been well documented and well covered in both the developed
and developing worlds includi
ng
Zimbabwe

for example O
y
ama
(
1997
)
investigated the
relationship between the stock market and macroeconomic variables in Zimbabwe using the
R
evised
D
iscount
M
odel
,
ECM and
M
ultifactor
R
eturn
M
odel, Sadosky
(2001) studied the
interaction between the stock
market and economic activity in the US
and Kurihara
(2006)
investigated the relationship between macroeconomic variables and daily stock pr
ices in Japan
the
period covered by the researcher is of great interest to many stakeholders who have
questioned the
ability
of macroeconomic
fundamentals
in explaining
ZSE given
the illicit
dealings that caused speculative behavior on the local bourse
.
In
addition,
given the fact that the year 20
07 was almost the dusk of the h
yperinflationary
environment and the use of
the Zimbabwean dollar as legal tender in Zimbabwe the researcher is
likely to close the curtain in the Zimbabwean dollar era
.
1.
5
R
esearch
Hypothesis
H
0
:
macroeconomic variables does not Granger cause ZSE index
H
1
:
macroeconomic variables Granger cause ZSE index
1.
6
R
esearch
Questions
In tackling this
research,
the researcher should answer the following questions:
Which of the selected macroeconomic variables has greatest influence on the performance of
the Zimbabwe
stock exchange?
Is there any causal relationship between the selected macroeconomic variables and the ZSE
industrial index?
If there is any causality what is the direction of the causality?
6
1.7 Assumptions
In carrying out this research the researcher
assumed the following:
All the information about the variables in question is authentic and satisfactorily reliable.
Therefore al
l the data sources used (CSO,
RBZ
and IMARA
) are reliable in collecting their
data.
The econometric package (E

views student’s
3.1 versions) used correctly computes the
relationship between the variables under consideration.
The causal ordering is unknown. It is unknown which variable is a cause and which one is a
result of the relation.
1.9
O
rganization
of
the
Study
This researc
h will be conducted in a system
atic
way as
follows:
Chapter two offer
s
both the theoretical and empirical
literature. Chapter
three considers the
methodology
employed by the researcher in the study. It also highlights the data used.
More so
,
it justifies t
he inclusion of certain variables
in the
study. Chapters four presents and analyze
the
findings from chapter three (results will obtained using E

views student’s
3.1
version
)
.
C
hapter
five summarizes the research findings and also proposing policy
recommendations drawn
directly from this study.
7
CHAPTER 2
LITERATURE REVIEW
2.0
Introduction
Sir Isaac Newton once said, “It is by standing on the shoulders of giants that you can manage to
see more than others.” The researcher therefore exploits the
ideas put forth by different
individuals about the particular issues in this chapter in order to add value to his research.
This chapter reviews both theoretical and experimental underpinnings of this study.
Literature on
the various studies on stock market performance is examined to give lessons to the current study.
Theoretical review analyses the theoretical premises of this while empirical assess past
researches for both developing and developed economies
that are relevant to this study.
There
are various opinions reflecting wide views towards the impact of real interest rates, exchange
rates and consumer price index on stock market performance.
The behavior of the world’s stock
market performances exhibit
diverse responses to the various macroeconomic variables in
question.
2.1
Theoretical Review
Irving Fisher (1930) found that real interest rates were equal to nominal interest rates minus
expected inflation. This macroeconomic relationship is known as th
e Fisher
Effect. The
Fisher
Effect is unique in that it incorporates expected inflation as opposed to actual inflation rates into
the analysis. This is crucial to this study because it allows the use of rational expectations model.
The Fisher Effect is pr
imarily an alternative way of measuring real interest rates and is used as a
means of relating interest rates and inflation expectations to stock prices. To fully understand the
relationship between the Fisher Effect and stock prices, it is necessary to un
derstand the
individual relationships between inflation expectations, interest rates, and the stock market.
An E
fficient
M
arket is market in which the values of securities at any instant in time fully reflect
all available information resulting in the mar
ket value and the intrinsic value being the same.
Propounded by Fama (1970) EMH also states that the prices of shares will immediately adjust to
8
new information. There are three different types of efficiency in an efficient market. These are
weak form, sem
i strong form and strong form.
The weak form fully incorporates information about past stock prices. Stock prices are said to
follow a random walk as the information on stock comes in a random manner. Attempts to use
technical analysis are futile in predic
ting future prices. In the semi strong form, prices
incorporate all publicly available information including published accounting statements as well
as historical price movements. Thus any information that can be extracted is already reflected in
the price
.
In the strong

form all information public or private is incorporated. It makes the most stringent
demand on information since it says that even the information available only to those closely
concerned with the firms has already been taken up and incorpo
rated in the price.
The main
fundamental implication of EMH is that if the markets are efficient then it is impossible for
investors to exploit information in order to earn excess returns over a sustained period of time
(Howells and Bain 2002).
As for the
effect of macroeconomic variables on the stock market EMH suggests that
competition among profit maximizing investors in an efficient market will ensure that all the
relevant information currently known about changes in macroeconomic variables are fully
r
eflected in current stock prices so that investors will not be able to earn abnormal profits
through prediction of future
prices, Chong
and
Goh
(2003)
.
The Capital Asset Pricing Model (CAPM) was developed by Sharpe and Litner in 1964.
It
is an
equation that equates the expected rate of return on a stock to the risk free rate and the risk
premium for the stock’s systematic
risk, Martin
and Scot jnr (1996). Basically, the CAPM
illustrates the relationship between expected return on an indiv
idual security and the beta of the
security. The beta according to
Hull (
1997) is a parameter that shows the relationship between
the return on a portfolio of stocks and the return on the market.
In all, the model argues that the investor requires excess r
eturns to compensate for any risk that is
correlated to the risk in the return from the stock market but requires no excess return for other
risks. The
Arbitrage Pricing Model (APM) was developed by Ross (1976). According to the
APM returns vary from their
expected amounts because of unanticipated changes in a number of
9
basic economic forces such as industrial production, inflation rates, term structure of interest
rates and the difference in interest rates between high and low risk bonds, Martin and Scot j
nr
(1996). It suggests that the risk of a security is reflected in its sensitivity to unexpected changes
in important economic forces.
The Classical case of investment and savings theory strongly argue that the savings are invested
as a result of interest
mechanism. They propel that in an economy where demand and supply
forces are at play, savings and investments are equated in the economy. They suggest that both
investments and savings are functions of interest rates. Savings are positively related to the
rate
of interest. As interest rates rise, people are induced to save more and the convex is typically
true. If on the other hand, the interest rates decline, depositors are discouraged to save and they
transfer their money from the money market to equity m
arket and the stock market will start to
perform well (Arnold, 1998).
On the other hand, investment is inversely related to interest rates. Therefore, as interest rates
rises, investments fall since th
e cost of borrowing rises also.
However under the
hype
rinflationary environment such as the case in Zimbabwe, people will rush to the equity
market where positive capital gains prevail. Hence most investors will rush to the stock market
where capital gains will be inevitable, Arnold (1998).
From the above sce
nario, it should come to light that low interest rates reflect themselves in
limited savings generation, especially in a hyperinflationary climate. In fact, investors will be
more concerned about real return that is nominal interest rates less inflation ra
te. Therefore, they
will discard the money market and invest in the stock market and property market where real
returns are positive. It should be clearly noted that investors are rationale and aim to maximize
capital gains, Arnold (1998).
2
.2 Empirical Re
view
Gunasekarage et al (2004) examined the influence of macroeconomic variables on stock market
equity values in Sri Lanka using the Colombo all share price index to represent the stock market
and money supply ,the Treasury bill rate (as a measure of interest
rates ),the consumer price
index as
(
a measure of inflation
)
and the exchange rates as the macroeconomic variables. With
10
monthly data for the 17

year period from January 1985 to December 2001 and employing the
usual battery test which included unit roots,
cointegration
, and the VECM, they examined long
run and short run relationships between the stock market index and economic variables. The
VECM analysis provided support for the argument that the lagged values of macroeconomic
variables have a significant
influence on the stock market.
Rigobon and Brian (2001) used an identification technique based on the heteroscedasticity of
stock market returns to identify the reaction of real interest rates to the stock market performance
in United States of America.
They found that real interest rates reacts significantly to stock
market movements, with a 5% rise or fall in the industrial index increasing the likelihood of a 25
basis point tightening (easing) by about half. The authors decompose both daily and weekly
movements in interest rates and stock prices from approximately 1985 to 1999. Their results
suggest that stock market movements have a significant impact on short

term interest rates,
driving them in the same direction as the change in stock prices. The au
thors attribute this
response to the anticipated reaction of monetary policy to stock market increases. They
acknowledge that this interpretation should be taken a bit cautiously.
Mukherjee and Naka (1995) applied Johansen’s (1988) VECM to analyze the rel
ationship
between the Japanese stock market and exchange rates , inflation , money supply ,real economic
activity ,long term government bond rate and call money rate. They concluded that a
cointegrating
relation indeed existed and that stock prices contrib
uted to this relation. Maysami
and
Goh
(2000) examined such relationship in Singapore. They found that inflation ,money
supply growth , changes in short and long term interest rate and variations in exchange formed
an cointegrating relation with changes in
Singapore stock market levels.
Vuyyuri (2005) investigated the cointegrating relationship and causality between the financial
and real sectors of the Indian economy using monthly observations from 1992 through December
2002. The financial variables used w
ere interest rates, inflation rate, exchange rate stock return
and the real sector was proxied by industrial productivity. Johansen (1988) multivariate
cointegration
test supported the long run equilibrium relationship between the financial sector
and the
real sector. Granger test showed unidirectional Granger causality between financial
sector and real sector of the economy.
11
Hardouvelis(1987), Keim(1985), Litzenberger and Ramaswamy (1982) empirically investigated
whether the main economic indicators such a
s inflation, interest rates ,TB returns, exchange
rates, and money supply effective explain the share returns in Turkey using the Johansen’s
(1988)
cointegration
analysis. They precisely concluded that there was a negative correlation
between interest rate
s and Standard and Poor’s index.
William Breen, Lawrence Glosten, and Ravi Jangannathan (1989) completed a study of the
relationship between the Treasury bill rate and the stock market in India using the usual battery
tests (VECM, unit roots tests
cointeg
ration
tests). The study found that an inverse relationship
between stock index returns and Treasury bill interest rates exists when a value

weighted stock
index is used. The reasoning behind this negative relationship is that, when interest rates rise, th
e
expected earnings streams decline because of the higher cost of borrowing and financing
expenditures. Because earnings reports play a dramatic role in stock prices, a rise in interest rates
that adversely affects earnings reports will lead to lower stock
prices (Breen et al, 1989). In
summary, the Fisher Effect should have a negative relationship with the S&P 500.
Masih and Masih (1996), Kwon et al (1997) Cheung and Ng (1998) and Nasseh and Strauss
(2000) examine the impact of several macroeconomic variab
les including real interest rates,
inflation rate, real economic activity, long term government bond rate and call money rate on
stock market performance in both developed and emerging economies. Most studies found that
real interest rate among other macro
economic variables have significant influence on the stock
market and/or the existence of a long

run relationship between these macroeconomic variables
and stock prices
.
Oyama (1997) looked closely at the relationship between stock prices and the macroecon
omic
variables in Zimbabwe using Revised Discount Model, Error Correction Model and Multifactor
Return Generation Model. This study used quarterly time series broad money supply(M2) ,three
month Treasury Bill rate and the IFC stock return index which consi
ders both capital and
dividends payments. The sample period was set from the first quarter of 1991 to the fourth
quarter of 1996. He noted that by using an error

correction model to stock returns and growth
rate of money and Treasury bill rates has been q
uite stable since 1991, except during the period
of partial capital account liberalization.
12
An analysis of individual stock returns indicates that the Zimbabwe Stock Exchange assimilates
changes in some important macro variables quite consistently. Still,
the contributions of these
macroeconomic variables cannot explain the volatile movements of stock returns during the
period from late 1993 to 1994.
In conclusion, Oyama
(1995)
noted that money supply (M2) and TB rate had an impact on the
performance of stock prices in Zimbabwe since 1991. It should be brought to light that many
theories, literature and empirical evidence strongly believes that the performance of stock market
ca
n be largely be affected by inflation rate, money supply growth, interest rate and drought.
Ibrahim (1999) also investigated the dynamic interactions between the KLSE composite index
and seven macroeconomic variables (industrial production index, money sup
ply M1 and M2,
consumer price index, foreign reserves, credit aggregates and exchange rate using the Johansen’s
(1988,1991,1992b) and Johansen and Juselius (1996
) multivariate cointe
gration techinique.
Observing that macroeconomic variables led the Malaysi
an stock indices he concluded that
Malaysian stock market was informationally

inefficient.
Chong and G
oh (2003) results were similar to the above they showed that stock prices, economic
activities, real interest rates and real money balances in Malaysia
were linked in the long run
both in the pre and post capital control sub periods.
Sadosky (2001) studied the interaction between the stock market and economic activity in the
United States using monthly data from 1974 to 1994. The macroeconomic variables c
onsidered
were the US production index, consumer price index and three month Treasury bill rate. The real
stock prices were calculated by deflating the nominal standard and poor index 500(S&P 500) by
the CPI. He used the Granger causality tests and noted t
hat causality runs from inflation to
changes in interest rates.
Interest rates on the other hand, predict changes in real stock returns
and changes in stock prices predict changes in industrial production.
He also concluded that there
was no evidence of Gr
anger causality running from real stock returns to inflation.
Maysami and Sims (2002, 2001a, 2001b) employed the Error Correction Modeling (ECM)
technique to examine the relationship between macroeconomic variables and the stock returns in
13
Hong Kong and S
ingapore (Maysami and Sims 2002b), Malaysia and Thailand (Maysami and
Sims 2001a), Japan and Korea (Maysami and Sims 2001b).
Through the employment of Hendry (1986) approach which allows making inferences in the
short run relationship between macroeconomic
variables as well as the long run adjustments to
equilibrium, they analyzed the influence of interest rates, inflation, money supply, exchange rate
and real activity along with a dummy variable to capture the impact of the 1997 Asian financial
crisis. The
results confirmed the influence of macroeconomic variables on the stock market
indices in each of the six countries under study though the type of magnitude of the association
differed depending on the country’s financial structure.
Kurihara (2006) choose
s the period March 2001

September 2005 to investigate the relationship
between macroeconomic variables and daily stock prices in Japan. He takes Japanese stock
prices, U.S. stock prices, exchange rate (yen/U.S. dollar), the Japanese interest rate etc. The
empirical results show that domestic interest rate does not influence Japanese stock prices.
However, the exchange rate and U.S. stock prices affect Japanese stock prices. Consequently, the
quantitative easing policy implemented in 2001 has influenced Japa
nese stock prices. Doong et
al. (2005) investigate the dynamic relationship between stocks and exchange rates for six Asian
countries (Indonesia, Malaysia, Philippines, South Korea, Thailand, and Taiwan) over the period
1989

2003. According to the study, t
hese financial variables are not cointegrated. The result of
Granger causality test shows that bidirectional causality can be detected in Indonesia, Korea,
Malaysia, and Thailand. Also, there is a significantly negative relation between the stock returns
a
nd the contemporaneous change in the exchange rates for all countries except Thailand.
Bhattacharya and Mukherjee (2003) investigated Indian markets using the data on stock prices
and macroeconomic aggregates in the foreign sector including exchange rates,
money supply as
well as interest rates and concluded that there is no significant relationship between stock prices
and exchange rates. In another study, Muhammad and Rasheed (2003) examined the relationship
between stock prices and exchange rates of four
South Asian countries namely Bangladesh,
India, Pakistan and Sri Lanka and found that there is no significant relationship between the
variables in India and Pakistan, either in the short

run or long

run. However, they found a bi

directional relationship
in the case of Bangladesh and Sri Lanka.
14
2
.3 Conclusion
Based on the afore mentioned literature, one can safely say that countable studies in Zimbabwe
and other emerging markets as well as developed nations have used macroeconomic
fundamentals to analyze
the performance of the stock market. Most studies used cross

country
data hence most focus was based on the difference in structures of economies and may explain
differences in economic performance across economies. The proceeding chapter focuses on the
me
thodological approach to be used in this study.
15
CHAPTER 3
METHODOLOGY
3.0
Introduction
This episode focuses on the method and the techniques to be used to come up with the solution to
the hypothesis. Furthermore the model to be employed for
testing the data is also explained. In
this chapter major aspects such as model specification, the estimation of the Sims (1972) basing
on the Granger multivariate model,
the unit root tests and cointe
gration tests are also discussed.
3.1
Model Specificat
ion
In this particular study, the researcher will make use of quantitative variables from time series
data. The data will be of quarterly set up. Analysis of the data will be through the econometric
model adopting Granger and Sims multivariate approach. Th
e researcher shall make use of the
econometric package of E

views 3.1 that will employ modern techniques in modeling of time
series data.
Although there is no universal definition of causation, Granger (1969) launched an approach
based on time series data
in order to determine causality. Causality in econometric sense refers to
the ability to predict. Thus causality in Granger (1969) and Sims (1972) sense: a variable x is
said to granger cause y if present y can be predicted by greater accuracy by using pas
t values of x
rather than not using such past values all other information being identical, from the definition of
causality if β
1
=β2=β3=0 the x does not granger cause y. However if any of the β coefficients are
non zero then x does granger cause y.
There
are different types of situations under which Granger causality test can be applied. These
include;
The simple bi
variate Granger causality where there are two variables and their
lags.
A multivariate Granger causality where more than two variables are cons
idered. This is
used where it is supported that more than one variable can influence the results.
Granger causality can also be tested in a Ve
ctor Autoregressive (VAR) frame
work where
a multivariate model is extended so as to test for simultaneity of all
included variables.
16
The researcher will employ Sims’ test basing on the Granger definition of causality.
A simple
regression equation which includes all the explanatory variables in this research can be stated
as:
Ind
t
= β
0
+ β
1
Ir
t
+ β
2
CPI
t
+β
3
Er
t
+ ε
t
……………………………………………3.0
Where ind
t
is industrial index in period t
Ir
t
is interest rate in period t
CPI
t
is consumer price index in period t
Er
t
is the Zimbabwean dollar to United States dollar exchange rate
The researcher chooses to
employ the Sims (1972) test, based on Granger’s (1969) definition of
causality. In Sims approach, a Granger causality relationship is expressed in two pairs of
regression equations by simply twisting independent and dependant variables as follows:
X
t
=β
1,1
X
t
–
1
+ β
1,2
X
t
–
2
+…+β
2,1
Y
t
–
1
+ β
2,2
Y
t

2
+
…
+
1,t
. …………(3.1)
Y
t
=β
2,1
Y
t
–
1
+ β
2,2
Y
t

2
+…+β
1,1
X
t
–
1
+ β
1,2
X
t
–
2
+…
…
+
2,t
………...(3.2)
X
t
=β
1,1
X
t
–
1
+ β
1,2
X
t
–
2
+…
…
+
1,t
………………………………….(3.3)
Y
t
=
2,1
Y
t
–
1
+
2,2
Y
t

2
+
…
+
2,t………………………………………………………
(3.4)
Where the X
t
( interest rates)
Consumer price index and the exchange
rate) which are the
exogenous variables and the Y
t
( industrial share index)which is the endogenous variable
Equations (3.1) and (3.2) are called unrestricted whilst (3.3) and (3.4) are restricted.
According to Granger’s definition of causality:
Y does not
cause X if and only if β
2,1
=β
2.2
=………=β
2,p
=0…….. (3.5)
X does not cause Y if and only if β
1,1
=β
1,2
=………=β
1,p
=0……. (3.6)
17
3.2
Diagnostic Tests
Diagnostic tests are conducted on a model in order to determine whether any assumptions of the
normal regression mo
del are violated. In this research the researcher will conduct the unit roots
and
cointegration
tests. However if there is any
cointegration
relationship it can
re

parameterised
as an
E
rror

C
orrection
M
odel. So if the variables in the equation are cointergrated the researcher
will as
a
matter of fact use ECM which will contain short and long run effects.
However if there
is cointergration there should Granger causality in at least one direction. Hence if
the researcher
find
that there is
cointergration
he will then test the direction of causality using Granger
causality.
3.2.1
Unit
Roots Tests
Most time series in economics demonstrates a trend over time. These time series are not
stationery implying th
at they don’t satisfy the requirements of weak stationery. Hence the unit
root test is viewed as the initial stage to determine
whether the
stock market (represented by the
industrial index) and the macroeconomic variables in question are a stationery proc
ess. The
researcher cannot
do away with unit root
tests since their existence can lead to spurious
regression. Spurious regression implies that relationship between variables can look statistically
significant but in actual fact there will be no meaningful relationship among the variables.
Basically there are various methods for testing unit roots and these include: Augmented

Dickey
Test (ADF), extension to the dickey fuller test for example Pantula tests, Phillips Peron
tests,
Kwaitowski

Phillips

Schmidt

shin (
KPS), Elliot

Rothenberg

stock
point
optimal (
ERS) as Ng

Perron tests. However in all the cases the idea is to search for data generating process(DGP) pure
random walk, random walk with a drift and random walk with a drift and time trend. The
researcher will implement the ADF tests
sinc
e they are relatively easy to understand and
compatible with E

Views 3.
3.2.2
Integration
and
Cointegration
Tests
The researcher will also conduct integration and
cointegration
tests.
Cointegration
can generally
be defined as an econometric concept which m
imics the existence of the long run equilibrium
relationship among econometric variables.
Cointegration
tests are important in determining the
presence and nature of equilibrium economic relationship.
18
3.2
.3
Causality
Tests
The researcher chooses to employ
the Sims (1972) test, based on Granger’s (1969) definition of
causality. In Sims approach, a Granger causality relationship is expressed in two pairs of
regression equations by simply twisting independent and dependant variables as follows:.
With Sims
test, the direction of causality is judged as follows:
The result of F test
Direction of causality
(a) (3.5) holds, (3.6) does not hold : X causes Y (X
天
(戩 (㌮㔩潥s n潴 h潬搬d(㌮㘩潬摳
†
: †† causes X
Y
堩
(c)⸠Neither (㌮㔩潲 (㌮㘩潬搠†††††† ††⁆ee摢dc欠扥tween X an搠Y
X
天
(搩 B潴h (㌮㔩n搠(㌮㘩潬搠††† ††⁘ an搠Yre in摥灥n摥nt
3.
3
Justification
of Variables
As alluded to above in the model specification, the researcher shall not use all of the variables
that can explain the performance of the stock market index but a few shall be used due to time
and cost of data gathering restrictions.
3.3.1
Ex
change
Rates
Exchange rates are hypothesized to have a positive relationship with the stock prices. Assuming
that the Marshal
–
Lerner condition holds , a depreciation in the Zimbabwean dollar will lead to
an increase in the demand for the Zimbabwean export
s their by increasing cash flows to the
country.
According to Makherjee and Naka (1995) if the local currency is expected to appreciate
the market will attract invest and through the contagion effect among macro economic variables,
the rise in demand will
push up the stock market level suggesting a positive correlation among
exchange rates and the stock market.
3.3.2
Consumer
Price Index (As
a
Measure
of
Inflation)
Previous studies by Fama and Schwert (1977), Jaffe and Mendelker (1976) pointed to a positive
relationship between inflation and stock prices. The researcher therefore hypothesized a positive
relationship when he said an increase in the rate of inflation is likely to lead to economic
tightening policies which in turn increase the nominal risk free
rate and hence the discount rate in
19
the discount valuation model. Ceteris paribus, inflation will move the stock market indices in the
opposite direction through the opposite movements of assets. Hence a negative impact of
inflation on the stock market is
expected. The researcher employs the quarterly inflation rates as
measured by the consumer price index.
3.3.3
Interest
Rates
Since the RBZ and the financial sector at large keep their information confidential, the researcher
faced an impossible task to ob
tain continuous flowing rate hence he employed the 90 day
Treasury bill rate obtained from the RBZ website. The Treasury bill rate has been employed by
many researchers among them Gunasekarage et al (2004) in their study in Sri Lanka. Basically,
the intere
st qualifies as a bench mark for all other interest rates and given the obvious fact that
interest rates move together in the same direction in an economy, the Treasury bill rate can be a
best proxy of interest rates movements. The Treasury bill rate is a
better estimate for the nominal
interest rates since it takes into consideration the opportunity cost of holding stock instead of
other short term money market instruments that are relatively liquid. This rate offers an
alternative destination for saving t
o investors who would compare the yield in the stock market
and the yield in the bond market.
Interest rates can result in a bullish market since low interest rates can lead to an outflow of
funds from the fixed to the variable yield market. Given that the
re is a positive relationship
between nominal interest rates and the risk free return of the valuation models, nominal interest
rates should move the stock market indices in an opposite direction. Hence a negative
relationship between Treasury bill rate an
d stock market indices.
3.3.4
The
Stochastic Error Term (The Disturbance Term)
Including this variable allows the model to be stochastic. This error term captures explicitly the
size of some errors or misses in our model. One can justify the error term on
the basis that it
measures inaccuracy of some measured variables. In addition it also captures human
indeterminacy. Lastly it can cater for the omission of innumerable chance events.
Summing up the arguments in the above discussion, the following signs a
re expected for the
specified model:
20
Stock Market Index = f (TBR, CPI, , ER)


+
3.4 Data
C
haracteristics
The study uses the industrial index as the dependent variables and
three
macroeconomic
variables as the independent variables for the Zimbabwean economy ranging from 199
5
(1) to
200
7
(4). The data is collected from various publications of the Reserve Bank of Zimba
bwe,
Central Statistical
Office, ZSE
and IMARA
. In an effort to li
mit the volatility of the data, the
study uses quarterly data. In years where there is only monthly data, an average figure is
calculated for each
quarter.
The
researcher used
Lisman and Sandee (1964) technique that was
also used for generating quarterly d
ata by Kereke( 1996.)
Variables to represent the industrial sector index, , interest rate, exchange rate
and
inflation are
respectively the
ind
, TB rate, ER (Z$X/US$1)
and
CPI. It should be noted that the exchange rate
is defined as Zimbabwean dollars (Z$)
per unit of the United States of America dollar (US$).
The USA dollar is used because most international transactions are quoted in USA dollars
.
3.5
S
trengths and
W
eaknesses of
t
he
M
odel
The model used [Sims(1972) multivariate model] is advantageous in t
hat it include many
variables that influence the stock prices hence increases the predicting power of the model.
However this model though it is multivariate in nature, excludes some variables which can be
important in influencing stock prices. These varia
bles can include political climate and policies
in the country, money supply etc.
3.5
Conclusion
Given the necessary information or data, it will then be possible to give the direction of causality
between the ZSE performance and the macroeconomic variable
s in question following the
procedures stated and provide outcomes that will help the formulation of policies that will
stimulate the resuscitation of the Zimbabwe economy. Results of this study are presented in
chapter 4
21
CHAPTER FOUR
DATA PRESENTATION
4.0
Introduction
Analysis and presentation of findings of the previous chapter are dealt with in this episode. The
results to be presented in this
chapter have
been obtained using E

Views 3.1 version. This
chapter will make use of tables to summarize the results. However full results are results are
presented in appendix
B.
4.1
.
0 Determining the Optimal Lag Lengthy: Akaike information criterion and Schwarz
crite
rion
Many researchers had tested causality up to the fourth lag. In addition most of them used the rule
of thumb that lags shoulb be at least one third to one quarter o the sample size , Gujarati (2005).
However the researcher used the Akaike and Schwarz c
riterions.
Table 1
: Optimal Lag Length
Lag
0
1
2
3
4
5
AIC
47.00549
41.51206
41.06884
37.226
30.55358
43.52028
SC
47.1579
41.81799
41.37771
37.53815
30.8812
43.83830
Using the Akaike information criterion and the Schwatz criterion the researcher would
use a
maximum of four (4) lags in testing the direction of causality.
This is because at four lags both
the AIC and th
e SC were at their lowest. Also it is after the fourth lag that both values began to
increase
4.1
Unit
Roots Tests and Cointegration Resu
lts
All the variables
were subjected to unit roots tests and the findings are pr
esented fully in
appendix B
:
The table below illustrates the results from unit roots tests:
22
Table
2
:
Results of Unit Root Tests
Variable
ADF statistic
Critical value
Level of
stationa
rity
I
n
dustrial index
9.363213
1.9473
I(0)
Consumer price index
16.93724

1.9473
I(0)
Interest rates

6.700062

1.9474
I(1)
Exchange rates
3.743794

1.9474
I(0)
Ε
t

6,40992

3.915304
*
I(0)
*The value was calculated using the McKinnon(1991)
method.
I(0) implies that the variable is stationery at level and hence it has no unit roots and I(1) indicates
that the variable
has one unit root and will become stationery after its first difference.
Since all the variables are not I(0)

the 90 day
T
reas
ury bill rate becoming stationery after the first
difference while other variables are stationery at their level, there is a possibility that the
variables are cointe
grated. Put it differently, there exist a long term equilibrium relationship
between the v
ariables in question. The
researcher runs
the OLS on the residual
(see appendix B
for detailed information)
. He then used the McKinnon’s
(1991)
method of calculating critical
values
.H
e then used
the calculated critical values
to compare with the ADF stati
stic from E

Views. For the residual to be stationery
be stationery, the ADF statistic should be greater than the
calculated level of
significance
. McKinnon’s
formula
can be stated as:
C(p)=
𝛟
∞
+
𝛟
1
(T

1
)+
𝛟
2
(T

2
)
Where p is the probability which is usually at
5
%, the
𝛟
s are
obtained from
the McKinnon
values and are found in the
tables
and n
is the sample size
.
The researcher calculated t
he critical
values as below:
C(5%)=(

3.7429)+(

8.352)(
50

1
)+(

13.41)(
50

2
)
=

3.9153004
9
In this research since 6.40901>

3.91530049
this implied that the residual is stationery
. Since the
residual is stationery this reinforced the notion that the
re
is a long run equilibrium relationship
between the variables.
If
there
is
a
cointergration
relationship
this would imply that there should
be Granger causality in at least one direction.
Gujarati (2005)
23
To enable the researcher to establish the causal link between the macroeconomic
variables and
the Zimbabwe stock exchange indus
trial index
, causality tests were employed and the results
were as f
ollows. The researcher tested
the direction of causality up
to
four
lags.
4.2 Results Presentation
H
0
: macroeconomic variables do not granger cause ZSE index
H
1
:macroeconomic variables
granger cause ZSE index
Table 3 Causality Results
Direction of
causality
Number of
lags
f

critical (5%)
f

calculated
Probability
Decision
ind→ir
1
2.76
3.60537
0.06374
Reject H
0
ir→ind
1
2.76
0.00037
0.98473
Accept H
0
cpi→ind
1
2.76
61.4928
4.4E

10
Reject H
0
ind↔cpi
1
2.76
0.24624
0.62205
Accept H
0
er→ind
1
2.76
18.9361
0.00000
Reject H
0
ind→er
1
2.76
6609.98
0.00000
Reject H
0
Ind→ir
2
2.76
2.21573
0.12110
Accept H
0
ir→ind
2
2.76
0.17016
0.84408
Accept H
0
cpi→ind
2
2.76
4925.29
0.00000
Reject H
0
ind→cpi
2
2.76
210.750
0.00000
Reject H
0
er→ind
2
2.76
10.6683
0.00017
Reject H
0
ind→er
2
2.76
3733.37
0.00000
Reject H
0
cpi→ind
3
2.84
310.027
0.00000
Reject H
0
ind→cpi
3
2.84
208.120
0.00000
Reject H
0
er→ind
3
2.84
1341.36
0.00000
Reject H
0
ind→er
3
2.84
315017.
0.00000
Reject H
0
ir→ind
3
2.84
64.1310
1.6E

15
Reject H
0
ind→ir
3
2.84
2.60785
0.06447
Accept H
0
cpi→ind
4
2.84
222.122
0.00000
Reject H
0
ind→cpi
4
2.84
16.9760
4.6E

08
Reject H
0
er→ind
4
2.84
190.085
0.00000
Reject H
0
ind→er
4
2.84
46115.8
0.00000
Reject H
0
ir→ind
4
2.84
5.35758
0.00160
Reject H
0
ind→ir
4
2.84
0.88734
0.48081
Accept H
0
To come up with the direction of causality, the researcher compared the F

critical at 5% with the
F

value. Alternatively this could be done
by comparing the F

value with the probability and one
will obtain the same value.
Table
4
:
D
irection of
C
ausality
variables
lag
Direction of causality
Industrial index and cpi
1
Uni

directional causality from
24
CPI to industrial index
2
Bi

directional
causality(ind↔cpi)
3
Bi

directional causality(ind↔cpi)
4
Bi

directional causality(ind↔cpi)
Industrial index and exchange
rate
1
Bi

directional causality(ind↔er)
2
Bi

directional causality(ind↔er)
3
Bi

directional causality(ind↔er)
4
Bidirectional causality(ind↔er)
Industrial index and interest
rates.
1
Unidirectional causality(ind→ir)
2
Bi

directional causality(ind↔ir)
3
Uni

directional causality(ir→ind)
4
Uni

directional
causality(ir→ind)
the
4.2
Interpretation
of Results
From the unit root tests of the variables and the residuals, it can be seen that the variables has a
long run equilibrium relationship since
they
are cointe
grated.
Industrial
index, consumer
price
index and exchange rates are stationery at level. This impl
ies that they have no unit roots.
Interests
rates on the other hand are intergrated of order one. This means that they
are
difference
stationery and become stationery after
differencing once.
In
addition since they are cointe
grated
this suggests that they
have
G
ranger causality at least in one direction, Gujarati (2005)
Tables 3 and 4 indicate
that the direction of
causality
depends critically on the number of lagged
terms included. If one takes consumer price index for example; when one lag is
included,
c
ausality
runs from C.P.I to ZSE index
(uni

directional causality)
but as the number of lags are
increased then
the is bi

directional causality
Wickremasinge (2006) shows reported that there is bidirectional causality between short term
interest rates and Un
ited States of America stock prices. However in this research the student
found a unidirectional causality between the short term interest rates and the Zimbabwean stock
market.
This maybe mainly due to the fact that the macroeconomic conditions in Zimbabwe
during the period under study was so volatile that they
cannot
be matched with those in the US
which were stable during the period under study. The causality results between in
terests rates
and the Zimbabwe stock exchange indicates that during the period under
study, information
about interest rates could be used to predict the stock prices. However it was impossible to use
information about stock prices to predict the nominal s
hort term interest rates.
Since there is a bi

directional
(feedback)
causality between industrial index, consumer price index
and exchange rates it implies that it was possible to predict say stock prices given the
information on exchange rates or consumer
price index and vice versa.
25
Clearly it can be seen that there is causality in at least one direction from the above scenarios.
This therefore indicates causation. Put it differently one can use information about one variable
to predict future movements in
another. Thus one can use information from consumer price
index to predict future movements in the ZSE index and vice versa.
4.3 Conclusion
Results from this research indicates that the industrial index, exchange rates, consumer price
index are integrated
of order zero[I(0)] and the 90 day treasury bill rate is integrated of order one
implying that it has one unit root.
Residual tests reveal
that the
residuals are stationery, another
strong indication that they are co integrated .causality tests indicate t
hat there is causality
between the stock market and the macroeconomic variables in question even though the direction
of causality differs with each variable.
CHAPTER FIVE
RECOMMENDATIONS
and CONCLUSIONS
5.0 Introduction
26
This chapter closes the study by giving a summary of research findings, pointing their
implication and also proposing policy recommendations
drawn directly from the study. It also
attempts to compare the objectives of the study with the study findings. Thi
s will enable the
researcher to establish whether the objectives were met. In addition
,
the limitation
s
of the study
are
given together with suggestion for
future
study.
5.1 Policy Recommendations
Causal
relationship identified in this particular study is
a crucial step towards giving information
to policy makers to enhance informed
formulation, effective
implementation and review of
macroeconomic policies.
Since there is a causal link between
C.P.I
and ZSE index,
it is recommended that
policies
targeting
inflation should not be independent from policy that regulates the activities of the
stock market. It
is
recommended that under the hyperinflationary environment such as those
that existed during most of the period under study, policies to reduce inflation
should be
compatible with the efforts to tame the
optimistic behavior
at the stock market which exert
upward pressure on inflation rate.
Since u
nfavorable interest
rates can drive the suppliers of loanable funds out of the money
market given that the sto
ck market is a safe haven. Usually
monetary
policies
which require
cutting down of interest rate
will be self defeating vis

à

vis stock market dealings.
The
researcher recommends that
monitory
authorities therefore should not cut the interests rates
but peg them in such a way that they act as an incentive for savings. In this way funds will
not be diverted to the stock market.
In this way the demand for stocks will be relatively low
hence not e
xerting pressure on the stock prices to rise.
Since stock markets and the exchanges rates have a causal
link, the
government should allow
the floating exchange rate to rule in the economy. Thus fixing the exchange rate can give
wrong information on the fut
ure price movements on the stock market. As a result due to the
operation of the fixed exchange rate regime in Zimbabwe in 2006 and 2007 it was difficult to
use the exchange rate to predict stock prices movements on the Zimbabwe stock exchange.
Since peggi
ng the exchange rate will lead to a wide variance between the black market rate
and the official exchange
rate, people
tend to buy and sell their foreign currency on the black
27
market. The proceeds from the black market will be used to buy stocks on the equ
ities
market. This will exert
pressure on
the share prices to rise. Thus the stock exchange would
end up
performing
superbly even under unfavorable macroeconomic conditions. Hence I
recommend the authorities to operate the freely floating exchange rate reg
ime.
5.2 Suggestions for Future Research
The study of causality in Zimbabwe should be extended to the period of the use of multiple
currencies and the period of economic recovery that is now characteristic of the Zimbabwean
economy. Future researchers
should compare whether there is any deviations and differences
in results obtained in the pre

dollarization
era and the post dollarization era.
In addition, future researchers should also analyze the short

run interaction between the
performance of the ZSE
and macroeconomic variable.
Most researchers looking at the causal
link
between the stock market performance and macroeconomic variables focuses mainly on
long run interactions. Therefore f
uture researchers can use tech
niques such as Variance
Decompositio
ns and Impulse response Functions to look at the short run relationships.
Furthermore future researchers can also study the dynamic interactions using daily data since
most researchers were using most notably annual, monthly and quarterly data. This will g
ive
more accurate results since daily figures do not deviate from each other greatly as compared
to annual or even monthly data.
More so
, future researchers should include as many variables as they can. Thus future
researchers can include even variables li
ke the legal and political
environment
since these
have great influence on the stock market performance.
5.4 Summary
This research looked at the causal relationship between macroeconomic variables and the
ZSE index. Chapter one looked at the introduction
to the study. It considered mainly the
background of the study,research hypothesis,statement of the problem as well as objectives
of the study. Chapter two looked at the empirical and theoretical literature view. It supported
the research with theoretical
and empirical postulations that supported the current study.
28
In chapter four the researcher looked at the methodology to be employed. chapter five looked
at the presentation of the data and analysis of it.
5.3
Conclusions
This study was carried out to
es
tablish the causal link between the ZSE and the macroeconomic
variables (exchange rates, consumer price index and short term nominal interest rates
. The period
of study stretches from 1995(i) to 2007(iii) and makes use of quarterly data.
Given the empirica
l findings a number of
conclusions
can be deduced. The existence of long run
relationship between the Zimbabwean stock exchange and macroeconomic variables indicates
the long

run predictability of the Zimbabwean equity prices. Put
it differently, movements
in the
ZSE equity prices are tied to the long

run movements in economic
fundamentals
. Since the
money market in Zimbabwe has negative real returns (average yield was far below inflation
rates) over the period under study the stock market remained as
a hed
ge against inflation.
Also
the causality
stock market and CPI is
crucial evidence that supports bullish behavior on the local
bourse. In addition stock market bubbles are evident.
REFERENCES
Adept Solutions, (2000), “
About Zimbabwe Stock Exchange
,”
ZSE Copyright.
Arnold
P (
1998)
U
sing
E
conomic
I
ndicators to
I
mprove
I
nvestment
A
nalysis
,
N
ew
York,
29
John Wiley and Sons.inc
Bhattacharya
T.S and Mukhejee (2003)
R
elationship
between
S
tock
P
rices and
Macroeconomic Aggregates:
Evidence from India
Cheng S and Ng E (1995)
A S
tudy of
W
hether
M
acroeconomic
F
actors
C
an
I
nfluence
S
tock
M
arket
S
hare
Returns: Global
E
conomy
Quarterly
Review
1(1):
315

335
Chong
C.S
and
G
oh
C
(2003)
L
inkages of
E
conomic
A
ctivity
S
tock
P
rices and
Monetary
Policy
:
The Case of
Malaysia
; Asian
E
conomic
R
eview 32:
260

261
.
Doong C,
H
ondroyiannis G and Papapetron (2005)
I
s there a
L
ong
R
un
R
elationship
B
etween
S
tock
R
eturns and
M
acroeconomic
V
ariables:
Evidence from
A
sia
F
ama
E.F
(1970)
S
tock
Returns, Real
Activity, Inflation
and
Money
:
American
economic
review 71,545

560
Fama E
.
F and
Schwert
W
.
A
(1977)
A
sset
R
eturns and
Inflation:
J
ournal of
F
inancial
.
Economics
volume 2: 200

212
Fisher, I (1930),
“
The Theory of Interest Rates
”,
Macmillan, New York
Jaffe and Mendelker G (1976)
the
F
ischer
E
ffect
F
or
R
isky
Assets;
A
n
E
mpirical
Investigation
:
Journal of Finance volume 31:447

456
30
Gujarati d. (1995)
Basic
Econometrics
, McGraw
Hill
, inc
Gunasekarange
A
.
P
isedtasasai.
A
and
P
ower
D
(2004)
M
acroeconomic
I
nfluence on the
Stock Market:
Evidence from Emerging
M
arkets in South Asia
Hardouvel
I.
S
(1987)
M
acroeconomic
I
nfluence on the Turkish
E
quities
Market
:
A
Journal Of Economics and Finance volume 25(1): 33

49
Hendry D (1986)
E
conometric
M
odeling with
C
ointergrated
V
ariables. An overview
:
Oxford Bulletin of Economics and Statistics volume 48(3) 200

212
Howells E and Bains T
(2002)
U
nderstanding
Markets
:an
I
ntroduction to the
T
heory,
Institutions And practice of markets Hertfodshire
Harvester Wheat
Sheaf
Hull
J
(1997)
O
ptions,
F
utures and
O
ther
Derivatives
(
3
rd
)
P
rentice
H
all
N
ew
D
elhi
I
ndia
Ibrahim
J
(1999)
D
ynamic
I
nteractions
B
etween the
C
omposite
I
ndex and
M
acroeconomic
Variables in
Malaysia
IMARA
Weekly Review (2005)
“Company News”
No. 3
.
Johansen
S
and
J
uselius
K
(1988)
M
aximum
L
ikelihood
E
stimation and
R
eference on
C
ointergration with an
A
pplication to the
D
emand for
M
oney
Oxford
B
ulletin of
E
conomics and
S
tatistics 52:169

210
Keim
R
(1985)
i
s there a
L
ong
R
un
R
elationship
B
etween
S
tock
R
eturns and
31
Macroeconomi
c
Variables: Evidence
from E
merging markets.
The journal of
Financial
Research
18 (
2):223

237
Litzenberger
D
and
R
amaswamy
(1982)
S
tock
R
eturns and
E
conomic
A
ctivity: the Turkish
Case: European
J
ournal of
F
inance
Kurihara (2006)
Macroeconomic
V
ariables and Japanese
S
tock
M
arket
.
Journal of
F
inance
And
Economics 32(2):417

425
Kwon
M
,
M
oradoglu
G
and
A
rgae
K
(1997)
the
R
elationship
between
the
S
tock
M
arket
returns And the Rates of Inflation in Malaysia.
Journal
of finance 38(2)1

34
Lisman G and
S
andee T (1964)
reducing the
volatility in time series data .time series
Analysis
Journal of economics and statistics
Martin and Scot jnr(
1996
)
T
rade
I
nvestment and the
E
nvironment
. London
E
arthscan
Masih
A
and
M
asih
J
(1996)
T
he
D
ecline of
I
nflation and
The
bull market of 1982

1996
.
A
Journal
of
F
inancial and
Q
uantitative
A
nalysis 37(2)
Maysami R.C and Sim H.H (2002)
M
acroeconomic
V
ariables and
T
heir
R
elationship with
Stock Returns: Error Correction Evidence from Hong Kong
and
Singapore the
Asia economi
c review
Maysami R.C and Sim H.H (2001a)
an
Empirical
I
nvestigation of the
D
ynamic
R
elations
Between
M
acroeconomic
V
ariables and the
S
tock
M
arket in
32
Malaysia and Thailand
Maysami R.C and Sim H.H (2001b)
M
acroeconomic
F
orces and
S
tock
R
eturns
:
A
G
eneral
ECM Analysis of the Japanese and Korean Markets
Muhammad A and Rasheed E (2003)
Stock Prices And Exchange Rates
:
Evidence From
South Asian Countries
Mukherjee T.K and Naka A (1995)
D
ynamic
R
elations between
M
acroeconomic
V
ariables
and The
Japanese Stock Market
:
an Application Vector Error
Correction model
Nasseh A and Strauss J (2000)
S
tock
P
rices,
D
omestic and
I
nternational
M
acroeconomic
A
ctivity:a
C
ointergration
A
pproach
:
Quarterly
R
eview of
Economics And Finance 40(2): 229

245
Oyama T
(1997)
Determinants of Stock Prices:
the Case of Zimbabwe
IMF Working Paper
RBZ
various publications
Rigobou
B
and
B
rian
A
(2001)
E
quity
M
arkets and
Monetary
P
olicy: evidence from the
Wall Street,
Washington DC
Ross S.A (1976)
the
A
rbitrage
T
heory of
C
apital
A
sset
Pricing
:
Journal Of
E
conomic
Theory
341

360
Smith.W,
C
ohn R A and
B
aume
B
(1994)
The
S
ecrets
Of
E
conomic
I
ndicators
,
N
ew
J
ersey
33
Wharton School Publishing.
Sadosky
T
(2001)
B
roken
T
rend
O
utput
in
M
odel of
S
tock
R
eturns and
E
conomic
A
ctivity
Applied financial economics volume
1
Sharpe,
S
teven A and Litner B
(1954)
S
tock
P
rices
E
xpected
R
eturns an
d
I
nflation
.
Journal
.
of Finance and Economics
Sims
C
(1972)
Causality Revisited
.
A Journal of Econometrica
.
Vuyyuri
S
(2005)
R
elationship
B
etween
R
eal and
F
inancial
V
ariables in
India:
a
Cointergration Analysis
www.zse.co.zw
APPENDI
X
Appendix A
Raw data
DATE
ER
CPI
IR
IND
Mar

95
8.44
11.50
68.72
2,942.03
Jun

95
8.55
11.76
68.68
3,614.10
34
Sep

95
8.80
12.73
55.50
3,808.38
Dec

95
9.31
13.29
61.24
3,972.62
Mar

96
9.59
14.22
13.40
4,828.69
Jun

96
9.86
14.40
15.52
5,404.42
Sep

96
10.31
15.36
18.17
6,992.15
Dec

96
10.82
15.49
27.63
8,786.26
Mar

97
11.24
17.06
31.71
10,169.97
Jun

97
11.36
17.48
30.70
10,437.40
Sep

97
12.45
17.56
30.25
9,804.02
Dec

97
17.69
18.57
26.03
7,196.43
Mar

98
16.17
21.83
43.50
8,795.71
Jun

98
18.01
22.69
58.34
7,417.93
Sep

98
31.41
23.13
55.17
5,730.73
Dec

98
37.31
27.23
79.75
6,408.40
Mar

99
38.17
33.34
106.53
8,975.30
Jun

99
38.01
35.23
162.85
9,829.47
Sep

99
38.16
39.25
162.85
11,825.41
Dec

99
38.14
42.72
133.00
14,426.64
Mar

00
38.15
50.28
95.08
14,759.84
Jun

00
38.19
56.11
150.00
15,200.66
Sep

00
52.53
63.58
55.50
19,196.83
Dec

00
55.06
66.31
61.24
17,984.33
Mar

01
55.09
80.50
13.40
29,197.63
Jun

01
55.07
94.80
15.52
39,484.24
Sep

01
55.04
121.70
18.17
47,714.14
Dec

01
55.04
144.60
27.63
45,351.89
Mar

02
55.04
171.80
31.71
48,090.75
Jun

02
55.04
203.50
30.70
77,232.98
Sep

02
55.04
292.00
30.25
99,520.98
Dec

02
55.04
432.30
26.03
103,495.09
Mar

03
532.58
563.40
43.50
179,530.62
Jun

03
826.45
945.10
58.34
277,301.90
Sep

03
826.45
1,622.30
55.17
648,932.85
Dec

03
826.45
3,019.90
79.75
401,542.93
Mar

04
4,314.50
3,852.00
106.53
347,708.43
Jun

04
5,346.90
4,674.10
162.85
693,147.07
Sep

04
5,615.20
5,702.90
162.85
871,123.53
Dec

04
5,712.65
7,028.70
133.00
1,097,492.53
Mar

05
6,082.06
8,616.90
95.08
2,483,961.01
Jun

05
9,899.14
12,354.20
150.00
2,856,530.00
Sep

05
26,003.66
26,224.60
265.00
6,176,377.29
Dec

05
84,587.57
48,205.60
340.00
18,483,883.97
Mar

06
99,201.58
87,340.90
420.00
31,045,930.90
Jun

06
100,916.00
420,000.00
510.00
54,873,198.61
Sep

06
259,576.00
1,125,000.00
66.30
388,686,440.00
Dec

06
258,920.00
2,800,000.00
66.30
569,864,000.00
Mar

07
259,116.00
3,475,000.00
112.50
4,026,437,840.00
Jun

07
255,549.00
5,300,000.00
280.00
43,133,619,830.00
Sep

07
30,705,000.00
7,983,000.10
365.00
86,494,551,710.00
35
Appendix B
UNIT ROOTS
B1 :AUGUMENTED DICKEY FULLER UNIT ROOTS TESTS ON CPI
ADF Test Statistic
16.93724
1% Critical Value*

2.6090
5% Critical Value

1.9473
10% Critical Value

1.6192
*MacKinnon critical values for rejection of hypothesis of a unit
root.
Augmented Dickey

Fuller Test Equation
Dependent Variable: D(CPI)
Method: Least Squares
Date: 04/14/10 Time: 05:23
Sample(adjusted): 1995:2 2007:3
Included observations: 50 after adjusting endpoints
Variable
Coefficien
t
Std. Error
t

Statistic
Prob.
CPI(

1)
0.498679
0.029443
16.93724
0.0000
R

squared
0.839944
Mean dependent var
159659.8
Adjusted R

squared
0.839944
S.D. dependent var
517563.4
S.E. of regression
207061.7
Akaike info criterion
27.33922
Sum squared resid
2.10E+12
Schwarz criterion
27.37746
Log likelihood

682.4805
Durbin

Watson stat
2.157633
36
B2 :AUGUMENTED DICKEY FULLER UNIT ROOTS TESTS ON EXCHANGE RATES
ADF Test Statistic
3.743794
1% Critical Value*

2.6090
5% Critical Value

1.9473
10% Critical Value

1.6192
*MacKinnon critical
values for rejection of hypothesis of a unit
root.
Augmented Dickey

Fuller Test Equation
Dependent Variable: D(ER)
Method: Least Squares
Date: 04/14/10 Time: 05:26
Sample(adjusted): 1995:2 2007:3
Included observations: 50 after
adjusting endpoints
Variable
Coefficien
t
Std. Error
t

Statistic
Prob.
ER(

1)
26.44134
7.062712
3.743794
0.0005
R

squared
0.206278
Mean dependent var
614099.8
Adjusted R

squared
0.206278
S.D. dependent var
4305531.
S.E. of regression
3835844.
Akaike info criterion
33.17747
Sum squared resid
7.21E+14
Schwarz criterion
33.21571
Log likelihood

828.4369
Durbin

Watson stat
1.324569
B3 :AUGUMENTED DICKEY FULLER UNIT ROOTS TESTS ON INDUSTRIAL
INDEX
ADF Test Statistic
9.363213
1% Critical Value*

2.6090
5% Critical Value

1.9473
10% Critical Value

1.6192
*MacKinnon critical values for rejection of hypothesis of a unit
root.
37
Augmented Dickey

Fuller Test Equation
Dependent Variable: D(IND)
Method: Least Squares
Date: 04/14/10 Time: 05:29
Sample(adjusted): 1995:2 2007:3
Included observations: 50 after adjusting endpoints
Variable
Coefficien
t
Std. Error
t

Statistic
Prob.
IND(

1)
1.081310
0.115485
9.363213
0.0000
R

squared
0.625076
Mean dependent var
1.73E+0
9
Adjusted R

squared
0.625076
S.D. dependent var
8.17E+0
9
S.E. of regression
5.00E+09
Akaike info criterion
47.52451
Sum squared resid
1.23E+21
Schwarz criterion
47.56275
Log likelihood

1187.113
Durbin

Watson stat
2.017351
B4 :AUGUMENTED DICKEY FULLER UNIT ROOTS TESTS ON INTERESTS RATES
(90 DAYTREASURY BILL RATE)
ADF Test Statistic

6.700062
1% Critical Value*

2.6100
5% Critical Value

1.9474
10% Critical Value

1.6193
*MacKinnon critical values for rejection of hypothesis of a unit
root.
Augmented Dickey

Fuller Test Equation
Dependent Variable: D(IR,2)
Method: Least Squares
Date: 04/14/10 Time: 23:19
Sample(adjusted): 1995:3 2007:3
Included
observations: 49 after adjusting endpoints
Variable
Coefficien
t
Std. Error
t

Statistic
Prob.
D(IR(

1))

0.978616
0.146061

6.700062
0.0000
R

squared
0.483134
Mean dependent var
1.735510
Adjusted R

squared
0.483134
S.D. dependent var
110.3791
S.E. of regression
79.35532
Akaike info criterion
11.60595
38
Sum squared resid
302268.8
Schwarz criterion
11.64455
Log likelihood

283.3457
Durbin

Watson stat
1.973128
B5:UNIT ROOT TESTS FOR RESIDUAL
ADF Test Statistic

6.409017
1% Critical Value*

2.6090
5% Critical Value

1.9473
10% Critical Value

1.6192
*MacKinnon critical values for rejection of hypothesis of a unit
root.
Augmented Dickey

Fuller Test Equation
Dependent Variable:
D(RESIDUAL)
Method: Least Squares
Date: 04/14/10 Time: 23:33
Sample(adjusted): 1995:2 2007:3
Included observations: 50 after adjusting endpoints
Variable
Coefficien
t
Std. Error
t

Statistic
Prob.
RESIDUAL(

1)

0.911848
0.142276

6.409017
0.0000
R

squared
0.456011
Mean dependent var

7211821.
Adjusted R

squared
0.456011
S.D. dependent var
4.97E+0
9
S.E. of regression
3.66E+09
Akaike info criterion
46.90025
Sum squared resid
6.57E+20
Schwarz criterion
46.93849
Log likelihood

1171.506
Durbin

Watson stat
1.974892
39
Appendix C
CAUSALITY TESTS
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:34
Sample: 1995:1 2007:3
Lags: 1
Null Hypothesis:
Obs
F

Statistic
Probability
CPI does not Granger Cause IND
50
61.4928
4.4E

10
IND does not Granger Cause CPI
0.24624
0.62205
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:36
Sample: 1995:1 2007:3
Lags: 1
Null Hypothesis:
Obs
F

Statistic
Probability
ER does not Granger Cause IND
50
18.9361
7.2E

05
IND does not Granger Cause ER
6609.98
0.00000
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:37
Sample: 1995:1 2007:3
40
Lags: 1
Null Hypothesis:
Obs
F

Statistic
Probability
IR does not Granger Cause IND
50
0.00037
0.98473
IND does not Granger Cause IR
3.60537
0.06374
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:38
Sample: 1995:1 2007:3
Lags: 2
Null Hypothesis:
Obs
F

Statistic
Probability
CPI does not Granger Cause IND
49
4925.29
0.00000
IND does not Granger Cause CPI
210.750
0.00000
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:39
Sample: 1995:1 2007:3
Lags: 2
Null Hypothesis:
Obs
F

Statistic
Probability
ER does not Granger Cause IND
49
10.6683
0.00017
IND does not Granger Cause ER
3733.37
0.00000
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:40
Sample: 1995:1 2007:3
Lags: 2
Null Hypothesis:
Obs
F

Statistic
Probability
IR does not Granger Cause IND
49
0.17016
0.84408
IND does not Granger Cause IR
2.21573
0.12110
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:42
Sample: 1995:1 2007:3
Lags: 3
Null Hypothesis:
Obs
F

Statistic
Probability
CPI does not Granger Cause IND
48
310.027
0.00000
IND does not Granger Cause CPI
208.120
0.00000
41
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:42
Sample: 1995:1 2007:3
Lags: 3
Null Hypothesis:
Obs
F

Statistic
Probability
ER does not Granger Cause IND
48
1341.36
0.00000
IND does not Granger Cause ER
315017.
0.00000
Pairwise Granger Causality Tests
Date: 04/20/10 Time: 00:45
Sample: 1995:1 2007:3
Lags: 3
Null Hypothesis:
Obs
F

Statistic
Probability
IR does not Granger Cause IND
48
64.1310
1.6E

15
IND does not Granger Cause IR
2.60785
0.06447
Pairwise Granger Causality Tests
Date: 04/18/10 Time: 11:09
Sample: 1995:1 2007:3
Lags: 4
Null Hypothesis:
Obs
F

Statistic
Probability
CPI does not Granger Cause IND
47
222.122
0.00000
IND does not Granger Cause CPI
16.9760
4.6E

08
Pairwise Granger Causality Tests
Date: 04/18/10 Time: 11:10
Sample: 1995:1 2007:3
Lags: 4
Null Hypothesis:
Obs
F

Statistic
Probability
IR does not Granger Cause IND
47
5.35758
0.00160
IND does not Granger Cause IR
0.88734
0.48081
Pairwise Granger Causality Tests
Date: 04/18/10 Time: 11:11
Sample: 1995:1 2007:3
Lags: 4
Null Hypothesis:
Obs
F

Statistic
Probability
ER does not Granger Cause IND
47
190.085
0.00000
IND does not Granger Cause ER
46115.8
0.00000
42
Comments 0
Log in to post a comment