Strathmore 
UNIVERSITY 
Modelling the Impact of Oil Prices on Stock Prices in 
Kenya 
Miano Charles Mwangi 
Student No. 072595 
Submitted in partial fulfilment of the requirements for the Degree of 
Bachelor of Business Science in Finance at Strathmore University. 
School of Finance and Applied Economics 
Strathmore University 
Nairobi, Kenya 
December, 2015 
Declaration 
I declare that this work has not been previously submitted and approved for the award of a 
degree by this or any other University. To the best of my knowledge and belief: the Research 
Project contains no material previously published or written by another person except where 
due reference is made in the Research Project itself. 
©No part of this Research Project may be reproduced without the permission of the author and 
Strathmore University 
Miano Charles Mwangi 
..... ~ ........................................... Signature 
....... \\ ~~- ... \J.~~ .. ~fD~~/. ..... ~9..\~ ....... Date 
This Research Project has been submitted for examination with my approval as the Supervisor. 
Ferdinand Othieno 
............ :: ............................................ Signature 
.L h. ~ .2..-ul::S: 
.. :1 .... ..J,-.!> ........ : ••••••••••••••••••••••••••••••••••••••••••••• Date 
School of Finance and Applied Economics 
Strathmore University 
ii 
Abstract 
The purpose of the study is to model the impact of oil prices on stock prices in Kenya using 
monthly data for between 2003 and 2015. The study uses the Johansen's multivariate 
cointegration test and the vector error correction model (VECM). The Johansen's cointegration 
test shows that the variables are cointegrated with at most one cointegrating vector and the 
cointegration estimate reveals that oil prices have a significant relationship with stock prices in 
the long-run in Kenya. The VECM model reveals that in the short-run, oil prices have a 
significant influence on stock prices. Similarly, in the long-run, the study finds that oil prices 
have a negative effect on stock prices in Kenya. To address the impact of oil price shocks on 
stock prices, the study uses impulse response and variance decomposition analysis. The 
impulse response results show that oil price shocks cause an immediate decline in stock prices. 
On the other hand, the cumulative effects of oil price shocks account for 9.02% ofthe variation 
in stock prices in the long-run. The study recommends policymakers, financial analysts and 
shareholders to take into consideration the effects of oil prices in their financial decisions given 
the significant impact of oil prices on stock prices in Kenya. 
Key words: Stock prices, Oil prices, Cointegration, Impulse Response, Variance 
decomposition 
iii 
Acknowledgments 
First and foremost, I thank God for providing me with good health, strength and the ability to 
complete this project. In a special way, I thank my parents for their unrelenting support, 
encouragement and love throughout the duration of my course. 
I am sincerely grateful to my supervisor, Mr. Ferdinand Othieno, for his guidance, wisdom, 
witty remarks and pleasant criticisms and for his continuous support especially during those 
difficult moments. 
To my friends who provided me with wonderful insights and motivation, God bless you 
abundantly. 
Finally to Strathmore University and the School of Finance and Applied Economics, I am 
grateful for giving me the opportunity to pursue my dream course. 
iv 
Table of Contents 
Declaration ................................................................................................................................. ii 
Abstract .................................................................................................................................... iii 
Acknowledgn1ents ..................................................................................................................... iv 
List of Figures .......................................................................................................................... vii 
List of Tables ........................................................................................................................... vii 
List of Abbreviations ............................................................................................................. viii 
Chapter 1 Introduction ............................................................................................................. 1 
1.1 Background Information ............................................................................................. 1 
1.2 Problen1 Statement ...................................................................................................... 2 
1.3 Research Objectives .................................................................................................... 3 
1.4 Research Questions ..................................................................................................... 3 
1.5 Significance of the Study ............................................................................................ 3 
Chapter 2 Literature Review .................................................................................................... 4 
2.1 Theoretical Framework ............................................................................................... 4 
2.2 Macroeconon1ic View ................................................................................................. 5 
2.2 Etnpirical Evidence ..................................................................................................... 6 
2.2.1 Short-run Effects .................................................................................................. 6 
2.2.2 Long-run Effects .................................................................................................. 7 
2.3 Knowledge Gap ......................................................................................................... 10 
Chapter 3 Research Methodology .......................................................................................... 12 
3.1 Research Design ........................................................................................................ 12 
3.2 Population and Sampling .......................................................................................... 12 
3.3 Data Collection .......................................................................................................... 13 
3.4 Data Analysis ............................................................................................................ 13 
3.4.1 Model Specification ............................................................................................... 13 
v 
3.4.2 Unit root tests and Lag Length Criteria ................................................................. 14 
3.4.3 Johansen Cointegration Technique ........................................................................ l5 
3.4.4 Vector Error Correction Model (VECM) .............................................................. 15 
3.4.5 Impulse Response Functions ................................................................................. 16 
3.4.6 Variance Decomposition ....................................................................................... 16 
3.5 Limitations of the Study ............................................................................................ 17 
Chapter 4 Findings ................................................................................................................. 18 
4.1 Descriptive Statistics ................................................................................................. 18 
4.2 Unit Root Test ........................................................................................................... 19 
4.3 Johansen Cointegration Test ..................................................................................... 19 
4.4 Effect of Oil Price on Stock Price ............................................................................. 20 
4.5 Vector Error Correction Model ................................................................................. 21 
4.6 Impulse Response of Stock Prices due to Oil Price Shocks ...................................... 22 
4.7 Variance Decomposition of Stock Prices .................................................................. 23 
Chapter 5 Discussion .............................................................................................................. 25 
5.1 Short-Run Effects of Oil Prices on Stock Prices ....................................................... 25 
5.2 Long-Run Effects of Oil Prices on Stock Prices ....................................................... 25 
5.3 Conclusion and Recommendations ........................................................................... 25 
References ................................................................................................................................ 2 7 
Appendices ............................................................................................................................... 31 
Figure 2: Impulse Response Functions over twelve month horizon ........................................ 31 
vi 
List of Figures 
Figure 1: Monthly NSE-20 Index values and ORB Prices from 2003 to 2015 ......................... 2 
Figure 2: Impulse Response Functions over twelve month horizon ........................................ 31 
List of Tables 
Table 1: Summary statistics: January 2003 to October 2015 .................................................. 18 
Table 2: Unit root test results ................................................................................................... 19 
Table 3: Summary of Johansen Co integration for the model variables ................................... 20 
Table 4: Long-run cointegration results among model variables ............................................ 20 
Table 5: VECM Esti111ate ......................................................................................................... 21 
Table 6: Wald Test result on short-run effects of oil prices on stock prices ........................... 22 
Table 7: Impulse Response of Stock Prices in Kenya due to oil price shocks over twelve month 
horizon ..................................................................................................................................... 23 
Table 8: Variance decomposition of stock prices in Kenya over twelve month horizon ........ 24 
vii 
List of Abbreviations 
ADF 
AIC 
CPI 
ECT 
FTSE 
GDP 
IRF 
KPSS 
NASI 
NSE 
NSE-20 
OECD 
OPEC 
ORB 
pp 
SBIC 
VAR 
VECM 
Augmented Dickey-Fuller 
Akaike Information Criterion 
Consumer Price Index 
Error Correction Term 
Financial Times Stock Exchange 
Gross Domestic Product 
Impulse Response Function 
Kwiatkowski-Phillips-Schmidt-Shin 
NSE All Share Index 
Nairobi Securities Exchange 
NSE 20-Share Index 
Organisation for Economic Co-operation and Development 
Organization of the Petroleum Exporting Countries 
Organization of the Petroleum Exporting Countries Reference Basket 
Philips-Perron 
Schwarz's Bayesian Information Criterion 
Vector Autoregressive 
Vector Error Conection Model 
viii 
Chapter 1 Introduction 
1.1 Background Information 
The recent drop in oil prices from USD 108.64 in July 2014 to USD 43.66 in October 2015 has 
reinvigorated the debate on the relationship between oil prices, financial markets and the 
economy. In addition, the widespread use of oil has increased dependency on this product 
within the supply-demand equilibrium and as a result, oil has gained a strategic position in the 
world (Kapusuzoglu, 2011). Given the importance of oil to the world economy, it has become 
crucial and to a cetiain extent pertinent to research on the stock market's ability to evaluate the 
real effects of events (such as oil shocks) that exogenously affect the economy (Jones & Kaul, 
1996). 
Oil price changes are assumed to affect stock price through the future cash flows of companies 
and through the discount rate (Basher, Haug, & Sadorsky, 2012). From a microeconomic 
standpoint, if oil prices rise and the firms are unable to fully pass this cost increase to their 
consumers, the firms' profits and dividends which are key drivers of stock prices will decline 
(Al-Fayoumi, 2009). The effect on stock price depends on the efficiency of the stock market. 
The overall impact of changes in oil prices on stock prices depend on whether a company is a 
consumer or producer of oil or oil related products (Basher & Sadorsky, 2006). Since there are 
more companies in the world that consume oil than produce oil, the overall impact for instance 
of rising oil prices on stock markets is expected to be negative. 
Given that Kenya is a net importer of oil, it implies that changes in the crude oil prices, in 
which Kenya has no role to play, always interferes with the dynamics of inflation (Misati, 
2013). Since a higher expected inflation raises the discount rate, a rise in oil price has a negative 
impact on stock market returns (Huang, 1996). These statements suppoti that oil price 
movements are an important and interesting topic to study because increases in oil prices are 
often indicative of inflationary pressure in the economy which in turn could indicate the future 
of interest rates and investments of all types (Sadorsky, 1999). 
In order to understand and contextualize the oil price-stock price interaction, it is important to 
determine whether the stock market is a potential channel through which oil shocks are 
transmitted to the Kenyan economy. Figure 1 below shows the trend of monthly NSE-20 values 
and ORB prices between the periods of 2003 to 2015. The lowest NSE-20 index value was 
1510.63 and the highest was 5774.27. Similarly, the lowest ORB price was 23.72 US dollars 
and the highest was 136.03 US dollars. An important observation from the two data series is 
1 
that both stock prices and oil prices were gradually increasing from 2003. However, there was 
a sharp decline in stock prices and oil prices in 2008 as a result of the global financial crisis. In 
addition, from mid-2014, there was a sharp decline in oil prices and from early 2015, stock 
prices have been gradually declining. 
5,000.00 
Figure 1: Monthly NSE-20 Index values and ORB Prices from 2003 to 2015. The graph shows a gradual 
increase in oil prices and share prices from 2003, then a sharp decline in both prices in mid-2008. In 2015, 
the graph shows a significant drop in oil prices and stock prices. 
1.2 Problem Statement 
Different studies have been conducted to empirically examine the effect of oil prices on stock 
prices in both developed countries and emerging markets, however their results can best be 
described as inconclusive. Studies by Le & Chang (2011), Narayan & Narayan (2010) and 
Masih et al. (20 11) find a positive relationship between oil price and stock price. On the other 
hand, studies by Sadorsky (1999), Park & Ratti (2008), Obemdorfer (2009) and Filis (2010) 
observed a significant negative relationship between oil price and stock returns. Nevertheless, 
Maghyereh (2004), Huang (1996) and Cong (2008) do not observe any relationship between 
oil price and stock market return. 
There are various studies covering the oil-stock price relationship in Africa (see for instance 
Effiong, 20 14; Hamma, 2014) and studies investigating the effects of oil prices on 
macroeconomic factors such as inflation particularly in Kenya (see for instance Misati, 2013), 
however in the Kenyan context there lacks a study that explicitly focuses on the effects of oil 
2 
pnces on stock pnces especially, the long-run effects. Therefore, this study attempts to 
investigate the nature of the oil-stock price relationship by determining the short-run and long-
run effects of oil prices on stock prices in Kenya. 
1.3 Research Objectives 
The study focuses on the following objectives: 
1. To determine whether there are sho1i-run effects of oil prices on stock prices in 
Kenya. 
u. To determine whether there are long-run effects of oil prices on stock prices in 
Kenya. 
1.4 Research Questions 
The study's research questions are: 
1. Do oil prices have short-run efiects on stock prices in Kenya? 
11. Do oil prices have long-run e±Iects on stock prices in Kenya? 
1.5 Significance of the Study 
This study is important as it aids in understanding the relationship between oil price and stock 
price which would provide relevant information for shareholders, foreign investors, financial 
analysts and portfolio managers on the ability of fluctuation in oil price in predicting stock 
returns. Similarly, understanding the relationship between oil price and stock market price is 
fundamental for an appropriate assessment of the impact of oil shocks on the performance of 
the stock market. 
Furthermore, through comprehending the linkage between oil prices and the stock exchange, 
the government can be able to take proactive measures against policies that are detrimental to 
the stock market and overall the real sector. Therefore, the government can determine proper 
strategies and timings of policy actions to ensure macroeconomic stability. 
3 
Chapter 2 Literature Review 
Since the late 1980s, there has been a greater focus and a broader perspective on the impact of 
oil prices on the financial markets especially on asset prices such as stock prices. Given that 
oil price changes is crucial for understanding the volatility in stock market prices, the literature 
suggests that there is no consensus among scholars and economists on its impact. To be precise, 
the studies have found oil-stock price relationship to be controversial and the evidence is 
mixed, with the focus mainly on developed and emerging markets economies. 
This chapter discusses the theoretical and empirical studies of the oil-stock price nexus that 
provides guidance to the rest of the study. The chapter is organised as follows. First, the 
theoretical underpinnings and macroeconomic view of the oil-stock price interaction. Second, 
a discussion ofthe empirical literature describing the short-run and long-run effects of oil prices 
on stock prices. Finally, the chapter concludes highlighting the knowledge gap that the study 
anticipates to fill. 
2.1 Theoretical Framework 
Theoretically, an asset price in a financial market is determined by its expected discounted cash 
flows (Huang, 1996). For stock prices, it equals the discounted values of expected future cash 
f1ows whose realized stock returns is given as: 
(1) 
where cis the cash f1ow stream; r is the discount rate; E(.) is the expectation operator; and pis 
the stock market price. From equation 1, stock market prices (or returns) are influenced 
systematically by two factors, namely, the expected cash f1ows and discount rates. 
According to Huang (1996), expected oil prices affect stock returns via the discount rate, which 
consists of both the expected int1ation rate and the expected real interest rate. Since, both the 
expected inflation and interest rates are influenced by oil price, for a net importer of oil, an oil 
price increase will put downward pressure on the foreign exchange rate and upward pressure 
on the expected domestic inflation rate. A higher expected inf1ation rate raises the discount 
rate, which has a negative impact on stock returns. Furthermore, oil price inf1uences the real 
interest rate as a higher oil price relative to the general price level leads to an increase in the 
real interest rate (Huang, 1996). This reduces stock prices following increased hurdle rate on 
4 
corporate investments (Effiong, 2014). Additionally, oil is considered an input in the 
production process. A rise in the oil prices raises the cost of production, which depresses 
aggregate stock price. 
Fmihermore, as countries urbanize and modernize their demand for oil increases significantly. 
Future oil demand is difficult to predict but is generally highly correlated with the growth in 
industrial production. Energy, financial markets and the economy are all explicitly linked 
together on a country's path of economic growth (Basher & Sadorsky, 2006). Hence, according 
to Basher & Sadorsky (2006), the net importers of fossil f-uels such as crude oil face a high risk 
from oil price changes and the impact on profits of companies in these countries is likely to 
play a large role in the development of these economies and their financial markets. 
2.2 Macroeconomic View 
The macroeconomic view is one of the five schools of thought that have a bearing on the stock 
price behaviour and focuses on indicators such as Gross Domestic Product (GDP), inf1ation 
and interest rates in understanding how the economy functions and also on the use of factor 
analysis technique to determine the factors affecting asset returns. The arbitrage capital asset 
pricing model (Ross, 1976) has been a primary motivation for earlier studies (see for instance 
Hamilton, 1983; Mork, Olsen, & Mysen, 1994). Among macroeconomic factors considered in 
models is crude oil price (see for instance Narayan & Narayan, 2010; Filis, 2010). This 
approach is based on the economic logic which suggests that everything does depend on 
everything else. Other approaches are: the fundamentalist approach, the technical approach, 
the efficient market approach and the random walk approach. 
However, from a macroeconomic view, an oil price hike acts like an inflation tax on consumers 
and with two consequences (Basher & Sadorsky, 2006). Firstly, consumers have less 
disposable income to spend on goods and services and have to find alternative energy sources. 
Secondly, non-oil producing countries have to bear the rising costs and are faced with 
increasing risk and uncetiainty caused by oil price volatility which negatively affects stock 
prices and reduces wealth and investment. On the contrary, an increase in oil price is expected 
to have a positive impact on stock markets in oil exporting countries, through income and 
wealth effects. This is due to a rise in government revenues and public expenditure on 
infrastructure and other mega projects (Al-Fayoumi, 2009). 
Oil price movement is an important barometer for investors to make necessary investment 
decisions and for policy makers to adopt appropriate policies in managing stock markets 
5 
(Adebiyi, 2009) . Stock markets are often seen as leading economic indicators. Stock market 
are an important means of channelling funds to the real sector , it is therefore pertinent to know 
the relationship between oil prices and stock markets in order for the government to take 
proactive measures against policies that are injurious to the stock market in particular and real 
sector in general (Nwosa, 2014). 
2.2 Empirical Evidence 
2.2.1 Short-run Effects 
The paper by Sadorsky (1999) investigates using an unrestricted V AR with GARCH effects 
with monthly data from 194 7:1 1996:4, the relationship between oil changes and real stock 
returns in the United States. The results confirm that oil prices and oil volatility have a 
significant impact on aggregate stock returns and through variance decomposition analysis, it 
is revealed that stock returns fall in the short term in response to a rise in oil prices. In addition, 
the study also revealed that oil prices have asymmetric effects with positive oil shocks having 
a greater impact on stock returns and economic activity than negative oil price shocks. 
Similarly, Papapetrou (2001) using a multivariate VAR approach examines the impact of oil 
prices on stock returns in Greece for the monthly period 1989:1 - 1996:6. Her variance 
decomposition analysis reveals that an oil price shock has a negative effect on stock returns 
especially for the first four months. 
In a slightly different study, Basher, 1-Iaug, & Sadorsky (2012) assess the importance of the 
relationship between stock prices and oil prices and the relationship between oil prices and 
exchange rates together into one empirical structural vector autoregression model with monthly 
data that covers from 1988:01 - 2008:12. The standard impulse response functions (IRFs) 
repmi that in the short-run, stock prices respond negatively to a positive oil price shock. 
Moreover, the results also indicate that increases in oil prices depress stock prices. Also, oil 
prices respond negatively to an unexpected increase in oil supply and oil prices respond 
positively to an unexpected increase in demand. 
In an examination on the short-run reaction of stock markets to oil price shocks in the Asia-
Pacific region, Nandha & Hammoudeh (2007) observed that stock markets in Phillipines and 
South Korea were oil sensitive when oil price is expressed in local currency only. In addition, 
the study observed that none of the countries studied was sensitive to oil prices expressed in 
US dollars, irrespective of conditions in the oil market. 
6 
Recently, Vinh (2014) investigated factors affecting Vietnam's stock prices including US stock 
price, foreign exchange rates, gold prices and crude oil prices.Using the daily data from 2005 
to 2012, the results indicate that Vietnam's stock prices are influenced by crude oil prices. 
Particularly, the study found that in the short-run, Vietnam's stock prices have a low con·elation 
with crude oil prices however Vietnam's stock prices indicated a high correlation with US 
dollar-VN Dong exchange rates and gold prices. 
In Kenya, Gatuhi & Mac haria (20 13) examined the relationship between oil prices, exchange 
rates, interest rates and other macroeconomic variables and the stock market. Using monthly 
data from January 2009 to December 2012 and applying the Pearson's bivariate correlation 
coefficient, the findings showed that diesel oil prices positively and significantly influences the 
performance of the Nairobi Securities Exchanges and similarly indicated that interest rates have 
a negative and significant influence on the performance of the Nairobi Securities Exchange. 
In Nigeria, Adaramola (20 12) examined the short-run and long-run dynamic effects of oil price 
on stock returns spanning over 1985:1 - 2009:4 using the Johansen cointegration tests. A bi-
variate model was specified and empirical results showed a significant positive stock return to 
oil price shock in the short-run. Contrastly, in the long-run, results showed a significant 
negative stock return to oil price shock. 
2.2.2 Long-run Effects 
In a study using daily data from 1979-1990 on stock pnces and futures pnces, the 
contemporaneous and lead-lag correlations between daily returns of oil futures contracts and 
stock return in the United States, Huang (1996) applied a multivariate vector autoregressive 
(V AR) approach to investigate possible lead, lag and feedback effects in the oil and stock 
markets. The results found no evidence of a relationship between oil prices and market indices 
such as the S&P 500. Based on the results, the study concludes that the much-touted influence 
of oil price shocks on the aggregate economy is more a myth than reality. 
Using quarterly data from 1947-1991 to determine whether the reaction of international stock 
markets (Canada, Japan, United Kingdom and United States) to oil' shocks can be justified by 
current and future changes in real cash flows and/ or changes in expected returns and using the 
Producer Price Index for fuels as a measure of oil prices, Jones & Kaul (1996) found that oil 
prices do have an effect on aggregate stock returns. 
In an analysis on the long-run relationship between the world pnce of crude oil and 
international stock markets using a cointegrated vector error correction model with additional 
7 
regressors and allowing for endogeneously identified breaks in the cointegration and error 
correction matrices, Miller & Ratti (2009) revealed evidence of breaks after 1980:5, 1988:1 
and 1999:1. The study also found a clear long-run relationship for the six Organisation for 
Economic Cooperation and Development (OECD) countries for the periods 1971:1 - 1980:5 
and 1988:2- 1999:9, suggesting that stock market indices respond negatively to increases in 
the oil price in the long-run. Nevertheless, the expected negative long-run relationship appeared 
to have disintegrated after 1999:9. 
By employing an international multi-factor model that allows for unconditional and conditional 
risk factors to investigate the relationship between oil price risk and 21 emerging economies 
covering the period 1992:12 - 2005:10, Basher & Sadorsky (2006) found that for daily and 
monthly data, oil price increases have a positive impact on excess stock market returns in 
emerging markets. However, for weekly and monthly data, oil price decreases have a positive 
and significant impact on stock market returns. 
Using a multivariate V AR analysis with linear and non-linear specifications of oil price shocks 
over the period 1986:1 2005:12, Park & Ratti (2008) observed a negative relationship 
between oil prices and stock returns in the United States and 12 European countries while for 
Norway, an oil exporting country, the study observed that the stock market responded 
positively to oil price rise. 
In relation to non-linear specifications of oil pnce, Ciner (200 1) exammes the causality 
between oil prices and stock return in the United States using daily data from 1979-1990 on 
stock prices and futures prices, and relying on nonlinear causality tests, he finds that oil shocks 
affect stock index returns and also proves that a significant nonlinear correlation exists between 
oil prices and stock returns. 
By applying a multivariate V AR model to examine the the impact of oil prices on real stock 
returns for Brazil, China, India and Russia from 1999:1 2009:9, Ono (20 11) found that real 
stock returns respond positively to some of the oil price indicators (US Produce Price Index 
and net oil price increase) with statistical significance for China, India and Russia however 
Brazil does not show any significance responses. Also, analysis of variance decomposition 
showed that the contribution of oil price shocks to volatility in real stock returns was relatively 
large and statistically significant for China and Russia. 
Using a bivariate GARCH model, Hamma (20 14) investigated the transmission of volatility 
and shocks between oil prices and seven sector Tunisian stock indices spanning 2 April 2006 
8 
to 12 July 2012. Empirical results indicated the existence of significant shock and volatility 
spillovers across oil and Tunisian sector stock markets, but the intensity of volatility 
interactions varied from sector to sector. In addition, the results showed that the spillover is 
unidirectional from oil markets to stock markets. CoiTespondingly, oil price shocks whether 
temporary or permanent can be difficult to distinguish and uncertainties related to large changes 
in oil prices can have significant effects on consumer confidence and therefore on growth 
(Masih, 2011 ). The impact of these oil price shocks is likely to be significantly greater in oil-
importing countries, especially where policy frameworks are weak, foreign exchange reserves 
are low and access to international capital markets is limited (Masih, 2011 ). Thus, a deeper re-
examination of the standard view concerning 'transient shocks', particularly for net importing 
countries is pertinent (Misati, 2013). 
In their investigation on the interactive relationships between oil price shocks and the Chinese 
stock market indices spanning from 1996-2007, Cong (2008) using multivariate VAR observed 
that oil price shocks have no statistically significant impact on stock returns on Chinese stock 
market indices. Also in another study, Maghyereh (2004) examined the relationships between 
oil price and stock market returns in 22 emerging economies for the period covering January 
1, 1998 to April 31, 2004. The study utilizing V AR analysis and found that oil price did not 
impact stock returns in those emerging economies. 
In his study, Nwosa (20 14) examined the relationship between oil prices and stock market price 
in Nigeria for the period spanning 1985:1 to 2010:4. The study utilized the Johansen's 
multivariate conintegration test and the vector error correction model (VECM). The Johansen's 
test showed that the variables are conintegrated and the cointegration equation revealed that oil 
prices have a significant relationship with stock market price in the long run. 
Correspondingly, Kapusuzoglu (20 11) investigated the long term relationships between 
National1 00, National 50 and National30 Index oflnstanbul Stock Exchange and international 
Brent oil price by applying the Johansen Cointegration test. The study determined that there 
was a long term relationship (cointegration) between each index and oil prices. Similarly, 
Narayan & Narayan (20 1 0) using the Johansen Cointegration test and evidence on oil prices, 
stock prices and exchange rates assessed the relationship between oil prices and Vietnam's 
stock prices with daily series from 2000 to 2008. The study found both oil prices and exchange 
rates, have a positve and statistically significant impact on Vietnam's stock prices in the long 
run and not in the short run. 
9 
In investigating the time-varying correlation between stock market prices and oil prices for oil-
importing and oil-exporting countries, Filis (2011) employed a DCC-GARCH-GJR approach 
with a monthly stock and oil prices from 1987-2009. Empirical findings showed that oil prices 
had a negative effect on all stock markets with an exception during the global financial crisis 
of 2008. An earlier study by Filis (201 0) examined the relationship among macroeconomic 
factors (consumer price index and industrial production), stock exchange and oil prices in 
Greece spanning from 1996:1 2008:6. He employed a multivariate V AR model and found 
that in the long-run, oil prices and the stock market exercised a positive effect on the Greek 
consumer price index, Additionally, oil prices were found to have a significant effect on the 
stock market and oil prices did not have any effects on industrial production. 
A different study was performed by Le & Chang (20 1 1) who examined the response of stock 
markets to oil price volatility in Japan, Singapore, South Korea and Malasia by applying the 
genaralized impulse response and variance decomposition analysis to monthly data spanning 
1986:01 - 2011:02. The results indicated that oil price shocks have non-linear impacts on stock 
market returns. Furthermore, the results suggested that the reaction of stock markets to oil price 
shocks varies across markets. Specifically, the stock market responded positively in Japan 
while negatively in Malaysia; the response in Singapore and South Korea was unclear. 
2.3 Knowledge Gap 
There is extensive literature that has investigated the impact of oil prices on stock prices 
(returns) albeit with mixed results. Conventional wisdom in literature is that oil price shocks 
lead to stock market decline. However, such impact may difier between oil-importing countries 
and oil-exporting countries depending on the relative importance of oil to their macro economy 
(Park & Ratti, 2008). Empirical studies have found different results regarding the oil-stock 
price relationship including, a lack of statistical significance in the positive relationship (see 
for instance Sadorsky, 1999; Filis, 201 0) or no relationship (see for instance Huang, 1996; 
Maghyereh, 2004 ). 
This oil-stock price relationship has been explored in stock markets of developed countries (see 
for instance Anoruo & Mustafa, 2007; Ciner, 2001). Emerging markets in Latin America and 
Asia have been covered (see for instance Ono, 2011; Basher & Sadorsky, 2006). Furthermore, 
various African states have also been covered (see for instance Adaramola, 2012; Hamma, 
2014 ). In the Kenyan context, the impact of oil price on various macroeconomic variables 
especially on inf1ation was covered by Misati (2013) within the periods 2008-2009 and 2011 
10 
and the study found that oil prices have a significant role on the overall inflation. However, 
there lacks a study that focuses on determining the long-run effect of oil prices on stock prices 
in Kenya. 
Taking an overview of these studies, it is obvious that the literature is inconclusive regarding 
the relationship between oil prices and stock prices, although the studies reviewed have tried 
to generate a clearer understanding of the oil-stock price relationship. There is therefore hardly 
any doubt that a possible relationship between oil price movement and stock prices could exist. 
A fundamental reason why it is difficult to reach a definitive conclusion regarding the link 
could be the intenelationships involved in establishing the relationship between oil prices and 
stock prices. Oil price movement can have a significant impact on the stock prices, but so can 
other factors such as economic growth and interest rates, that are related to the stock market 
which the study addresses by including those factors in the study's model and hence 
highlighting the knowledge gap. Thus, whether these relationships exist in Kenya is the focus 
ofthis study. 
11 
Chapter 3 Research Methodology 
3.1 Research Design 
According to Parahoo (2006), the design selected should be the one most suited to achieve a 
solution to the proposed research questions. Following this study's research questions, the 
researcher employs an explanatory research design in applying standard methods discussed in 
the literature review. 
This study utilizes quantitative methods. The data analyzed is monthly from January 2003 to 
October 2015. That time period was chosen for the study since it covers the scope which 
exhibited strong volatility in international oil prices especially during the global financial crisis 
in 2008-2009 and the recent sharp decline in the world oil prices in the mid-20 14, early 2015. 
The study employs the Johansen's multivariate cointegration test and the vector eiTOr 
correction model (VECM) in its analysis. 
3.2 Population and Sampling 
The population of interest in this study is fi:om the Nairobi Securities Exchange (NSE) 
particularly, the daily values of the NSE 20 Share Index (NSE-20) and the OPEC Reference 
Basket (ORB) monthly spot oil prices from January 2003 to October 2015. Given that the study 
uses monthly prices, the sample that represents these daily values in the analysis is calculated 
for every last trading day in each month for the NSE- 20. Based on Fama's (1981) hypothesis 
that measures of economic activity have played a role in the analysis of stock market activity, 
the study also considers short term interest rate measured by the deposit interest rate and 
economic growth measured by the gross domestic product deflated by the GDP deflator, which 
may influence the relationships between oil prices and stock prices. 
The study analyzes monthly data since this was the most popular frequency in the literature 
(Sadorsky, 1999; Papapetrou, 2001; Park & Ratti, 2008; Le & Chang, 2011; Ono, 2014). Using 
data of a lower frequency such as quarterly is likely to introduce smoothing effects, thus, 
wiping out out any meaningful results (French, 1987). Daily data would pose problems 
especially because of the complexities of dealing with weekends and public holidays. 
The NSE-20 is selected since it is a good proxy for the whole stock market given that those 20 
proxy companies represent the largest market capitalization with a 75% market capitalization 
of the entire market. On the other hand, other indices such as the NASI that includes stocks 
which are not actively traded especially due to lack of liquidity and given that it was only 
12 
introduced in 2008 would therefore be inappropriate as the proxy. Moreover, using FTSE 
indices would be problematic given that the data concerning the index components are kept 
confidential for commercial purposes. 
ORB is chosen as a representative of the world oil prices since it is a weighted average of prices 
for petroleum blends produced by OPEC countries and ORB particularly incorporates Murban 
which is the crude oil imported by Kenya from United Arab Emirates. Neve1iheless, there are 
prices of other types of oil, namely, West Texas Intermediate, Brent and Dubai that can serve 
as a benchmark for other types of crude oil, however the choice of crude oil prices would not 
significantly affect the study since crude oil prices have been observed to f1uctuate in the same 
direction empirically (Chang & Wong, 2003). Furthermore, the study does not use domestic 
oil prices since they are inf1uenced by several factors such as price controls, high and varying 
taxes on petroleum products, exchange rate f1uctuation and domestic price index variations. 
Taking into account such considerations, the use of world oil price in US dollars and conve1iing 
into domestic currency through the market exchange rate is justified (Le & Chang, 2011 ). 
3.3 Data Collection 
All data used in this study is from secondary sources. The data series comprises of oil prices, 
stock prices, short term interest rates and economic growth. Oil prices are acquired from the 
OPEC database. Stock prices are obtained from the NSE database. Short term interest rates and 
economic growth indicators are sourced from the International Financial Statistics ofthe World 
Bank. 
Considering the inflation factor, oil price and stock price are transformed into real terms. Oil 
prices are converted from US dollars into the domestic currency and then deflated by the 
domestic price index for inflation-adjusted real values. Stock prices are deflated for real values 
by the domestic Consumer Price Index (CPI). 
For modeling purposes, all the data series are transformed to logarithmic form to stabilize the 
variability in the data (Narayan & Narayan, 2010; Kapusuzoglu, 2011; Le & Chang, 2011; 
Nwosa, 2014) with exception to interest rate (Nwosa, 2014). 
3.4 Data Analysis 
3.4.1 Model Specification 
The model is summarized as follows: 
(2) 
13 
where SP is stock price, OILP is oil price, itr is short term interest rate and RGDP is economic 
growth. 
Equation 2 can be written in the form of a multivariate VAR model as: 
Equation 3 can be written more compactly as: 
k 
xt = ao + {31 L xt=j + Et 
j=1 
(3) 
(4) 
where Xt is a 4x }-dimensional vector of endogenous variables of the model, a 0 is a 4x }-
dimensional vector of constant and ~1 is a 4x4 dimensional autoregressive coefficient matrices 
of established parameter and Et is k-dimensional vector of the stochastic enor term normally 
distributed with white noise propetiies N (0, 0"2). 
3.4.2 Unit root tests and Lag Length Criteria 
Various parametric and non-parametric tests have been developed for finding whether a series 
is stationary or non-stationary (i.e., includes a unit root). To ascertain the order of integration 
of the variables, the researcher is testing for unit root using, namely, Augmented Dickey-Fuller 
(ADF) and Phillips & Penon (PP) (1988). 
Null hypothesis for both tests shows that a unit root exists in the autoregressive representation 
of the time series while alternative hypothesis shows that the series is stationary, If the 
calculated value is greater than the absolute critical value, then null hypothesis is rejected and 
series is said to be stationary. 
Optimal Jag lengths are selected by using the Schwarz's Bayesian information criterion (SBIC) 
rather than using the Akaike information criterion (AIC) since SBIC embodies a much stiffer 
penalty than AIC. However, the excess sensitivity of the results obtained from ADF and PP 
tests to the lag length determined has been criticized time to time. Therefore, Kwiatkowski 
(1992) developed KPSS test, which is not sensitive to lag length. Null hypothesis of KPSS 
stationary test is the reverse of the null hypothesiss of ADF and PP unit root tests. This means 
that for the KPSS test, the null hypothesis is that the unit root does not exist (stationary) and 
alternative hypothesis signifies that time series is non-stationary (unit root exists). 
14 
3.4.3 Johansen Cointegration Technique 
Cointegration is an important technique to examine whether the economic and financial time 
series are co integrated (Vinh, 2014 ). According to Granger & Weiss (1983 ), cointegration is 
the statistical equivalence of the economic theoretical notion of stable long-run relationship. It 
is based on the properties of the residuals from the regression analysis when the individual 
analysis are non-stationary. 
To test for a long-run relationship among the variables, they must have the same order of 
integration (Engle & Granger, 1987). The multivariate cointegration technique of Johansen & 
Juselius (1990) is employed to test the relationship. The Johansen approach to cointegration is 
based on a VAR model. Johansen (1988) suggests two tests, namely the trace and maximum 
eigenvalue test statistic. 
The cointegration tests are based on the following equation: 
where fl. is the first-difference operator, k denotes lag length, Xt is a vector of m non-stationary 
endogenous (stock price) variables, n and TI are a matrix of coefficients, !l is a vector of 
constants, and St is a vector for errors. The lag length, k, is chosen to ensure that the errors are 
independent and identically distributed. 
The trace test examines the null hypothesis that the number of cointegrating vectors in the 
system, r, is less than or equal to ro where ro < p and pis the number of variables in the system, 
whereas the alternative hypothesis is that the impact matrix is of full rank. 
The maximum eigenvalue test examines the null hypothesis that there are r0 cointegrating 
vectors versus the alternative of ro + 1 cointegrating vectors. The results imply the existence 
of a co integrating relationship between oil prices and stock prices. 
3.4.4 Vector Error Correction Model (VECM) 
If the variables are integrated of the same order and are cointegrated, then the V AR model must 
include an error correction term (ECT) (Engle & Granger, 1987). Hence the V AR model 
incorporating the error correction model (ECM) is specified in a VECM as: 
k 
!J.Xt-1 = ao + TrXt-l + I rj !J.Xt-j + Ej 
j=l 
15 
(6) 
where /::,. is the difference operator, Xt-1 is a 4x !-dimensional vector of non-stationary I( 1) 
endogenous variables of the model, a0 is a 4x }-dimensional vector of constant and Et is k-
dimensional vector of the stochastic error tenn normally distributed with white noise properties 
N (O,o-2). n is the long-run matrix that determines the number of cointegrating vectors that 
consist of a and f3 representing speed of adjustment towards long-run equilibrium and long-
run parameter. r is the vector of parameters that represents the short-term relationship. VECM 
model of equation 6 enables the determination of the direction of causation between observed 
variables while providing estimate on both the long-run and short-run (Nwosa, 2014). 
The VECM model differentiates between the short and long-run dynamic relationships and 
tests for the hypothesis that the coefficients of lagged variables and the error correction terms 
calculated from the cointegrating regression are zero. If the coefficients in the system are jointly 
significant, then the lagged variables in the system are important in predicting current 
movements of the dependent variables (i.e., the shoti run dynamics) and the dependent 
variables in equation 6 adjust to the previous period's equilibrium error (Anoruo & Mustafa, 
2007). Because the cointegrating vectors bind the long-run behaviour of the variables, the 
VECM is expected to produce results in the impulse response analysis and variance 
decomposition that accurately reflects the relationship among the variables. 
3.4.5 Impulse Response Functions 
The direction, timing and duration of stock market prices response to each oil price shocks is 
analysed using impulse response functions (IRFs). IRFs are often used to explain the dynamic 
effects of the shocks on the endogenous variables. An IRF traces the effect of a one-time shock 
to one of the innovations on current and future values of the endogenous variables and analyses 
the interactive responses between oil price and stock price and obtains the contribution of oil 
price shocks to the variability in stock returns. 
3.4.6 Variance Decomposition 
Since estimated coeHicients from V AR models often appear to be lacking in statistical 
significance due to the inaccuracy of the technique in estimating standard errors, IRFs and 
forecasting variance decomposition are often used to explain the dynamic effects of the shocks 
on the endogenous variables. 
16 
While IRFs trace the effects of a shock to one endogenous variable on to the other variables in 
the V AR, variance decomposition separates the variation in an endogenous variable into the 
component shocks to the V AR. Thus, the variance decomposition provides information about 
the relative importance of each random innovation in affecting the variables in the V AR. 
Pesaran & Shin ( 1998) propose a more general alternative to the Cho leski decomposition which 
is unaffected by the ordering of the variables and which does not require the orthogonalisation 
of the reduced form innovations. The resulting responses are unique and fully take account of 
the historical patterns of correlations observed amongst the different shocks. 
3.5 Limitations of the Study 
This study makes assumptions especially regarding the NSE-20 index. The NSE-20 index's 
constituents have been regularly revised within the scope of the study. For example, Uchumi 
Supermarket Limited and Mumias Sugar Limited were replaced by Centum Investments 
Limited and CFC Stanbic Holdings Limited in 2014 while CMC Holdings Limited was 
suspended from trading in 2011. This poses a problem since the data is not as consistent as the 
study might have required. However, the study assumes away such change as insignificant to 
the findings. 
In addition, the study does not capture the historical dynamics of regime changes and also does 
not take into consideration structural breaks occurring within the scope of the study given the 
volatiltiy of the stock market and oil prices. Rather the study relies on Johansen cointegration 
test whose findings have been found under some circumstances to be consistent with structural 
break cointegration tests. 
Furthermore, the study examines a bivariate model involving oil prices and stock prices as the 
main variables rather than augmenting the bivariate model to a multivariate model by including 
other main variables such as exchange rate. This can limit the study since it is argured in 
economics literature, that bivariate models give rise to omitted variable bias (Narayan & 
Narayan, 201 0). 
17 
Chapter 4 Findings 
This study examines the short-run and long-run effects of oil prices on stock prices in Kenya. 
The study adopts the VECM model to differentiate between the short and long-run dynamic 
oil-stock price relationship. Similarly, the study adopts to use the impulse response function 
and variance decomposition to explain the effects of oil shocks on stock prices. 
4.1 Descriptive Statistics 
Table 1: Summary statistics: January 2003 to October 2015 
LSP LOILP 
Mean 8.2722 8.8403 
Median 8.1773 8.9030 
Maximum 8.8875 9.6096 
Minimum 7.7775 7.9162 
Standard Deviation 0.2910 0.3229 
Skewness 0.4107 -0.6894 
Kurtosis 2.1201 3.5102 
J arq ue-Bera 9.2957 13.8681 
Probability 0.0096 0.0010 
Observations 154 154 
Note: LSP (logarithm ofNSE-20 index); and LOILP (logarithm of oil prices). 
Table 1 displays the results of the preliminary data analysis for the model's variables: monthly 
oil prices and NSE-20 index considered in the scope of the study. The mean return for both 
variables are positive, with maximum and minimum returns almost equal in magnitude. 
Regarding whether the series are distributed normally or not; skewness, kurtosis and Jarque-
Bera statistics were considered. Both variables are not normally distributed; rather the oil price 
variable is negatively skewed and has a kurtosis in excess of 3 indicating it has a leptokurtic 
distribution while NSE-20 index is positively skewed and has a kurtosis of less than 3. The 
Jarque-Bera statistic and corresponding p-value confirm the rejection of the null hypothesis of 
normality at the 1% significance level. 
18 
4.2 Unit Root Test 
Table 2: Unit root test results of the model variables. The variables are shown to be 
stationary at the first difference by the ADF test. 
Augmented Dickey-Fuller test Phillip-Penon test 
Variables Level 1st Diff Status Level 1st Diff Status 
loilp -1.5967 -9.6415 I(l) -1.3873 -9.7212 I(l) 
~p -2.7360 -11.1555 I(l) -2.8703 -11.3693 I(l) 
lrgdp -1.0031 -5.7015 I(l) -1.4923 -2.6543 I(1) 
rtr -3.5328 -4.3548 I (I) -3.2293 -13.0793 I(l) 
Asymptotic critical values 
1% -4.0192 -4.0469 -4.0192 -4.0196 
5% -3.4395 -3.4528 -3.4395 -3.4397 
KPSS test 
Level 1st Diff Status 
0.2332 0.0527 I(l) 
0.1612 0.0980 I(l) 
4.1002 0.0907 I(1) 
0.0661 0.0450 I(l) 
0.2160 0.2160 
0.1460 0.1460 
Note: loilp (log of oil prices );lsp (log of stock prices );lrgdp (log of real gross domestic product) 
and itr (short-term interest rate). The optimal lag length was determined through SBIC. 
Table 2 shows that all variables are non-stationary according to ADF and PP tests at the 1% 
and 5% level of significance in their level form except for short-term interest rate which is 
stationary at level according to the KPSS test. All variables are stationary at the first difference, 
that is, integrated of order one I(l) by the ADF unit root test. Findings obtained from PP test 
are consistent with results of ADF test except for the coni1icting evidence by the logarithm of 
real GDP which is not integrated from the first degree I( 1) using the PP test. From these results, 
it will be possible to look into the matter to whether there is a long-term relationship 
(co integration) between oil prices and stock prices. 
4.3 Johansen Cointegration Test 
Having confirmed the stationary status of the variables, the study proceeded to examine the 
long-run relationship (co integration) among the model variables. The results are presented in 
Table 3. 
19 
Table 3: Summary of Johansen Cointegration for the model variables. The table shows 
the existence of at most one cointegrating vector for the model variables. 
Ho HA Eigenvalue leT race 95% lcmax 95% 
r=O r=1 0.2066 54.9848* 47.8561 32.4127* 27.5843 
r::S1 r=2 0.0827 22.5720 29.7970 12.0865 21.1316 
r::S2 r=3 0.0439 10.4855 15.4947 6.2942 14.2646 
r<3 r=4 0.0294 4.1913 3.8414 4.1913 3.8414 
Note: r indicates the number of cointegrating vector. Critical values are from MacKinnon 
(1999) P-values. * indicates significance of the test statistic at 5% level. 
The trace statistics and the maximum eigenvalue statistics reveal that the null hypothesis of no 
cointegration, for r=O was rejected. The stated values of these tests were greater than their 
critical values. On the other hand, the null hypothesis of no co integration, for r::S1 could not be 
rejected by both the trace and maximum eigenvalue statistics. The statistical values of these 
tests at r::S1 were less than their critical values which show evidence of a long-run relationship 
with at most one cointegrating vector that suggest that the variables are jointly determined. 
4.4 Effect of Oil Price on Stock Price 
The resulting long-run relationship between the measures of oil prices and stock prices are 
estimated using VECM model of equation (6), presented in a long-run cointegration in table 4 
below: 
Table 4: Long-run cointegration results among model variables. Table shows a negative 
change in stock prices due to change in oil prices and real gross domestic product. 
Cointegrating Coefficient LOILP ITR LRGDP 
LSP 18.2263 -0.8967 0.0679 -0.2944 
Standard Error - 0.3027 0.0457 0.1445 
t -statistic - -2.9628 1.4877 -2.0367 
Note: LOILP (Log of oil prices); LSP (Log of stock prices); ITR (Short-term interest rate) and 
LRGDP (log of real gross domestic product). 
Table 4 reveals that oil prices (loilp ), real gross domestic product (lrgdp) and short term interest 
rates (itr) are significant determinants of stock prices in Kenya. In the long run, the result 
implies that there is 0.897 percent and 0.294 percent negative change in stock market price due 
to a 1 percent change in oil price and real gross domestic product, respectively. However, if 
there is a 1% increase in the short-term interest rate, it would increase the stock market price 
by 0.068 percent in the long run. Based on the Johansen cointegration test and result of the 
20 
cointegrating equation, the study surmises that the long-run effect between oil prices and stock 
prices is negative. This result is similar to that observed by Miller & Ratti (2009) on the 
negative long-run effect of oil prices on stock prices. 
4.5 Vector Error Correction Model 
Based on the cointegration test, VECM captures the short-run and long-run interaction between 
oil prices and stock prices, presented in Table 5 below. 
Table 5: VECM Estimate. From the table the error conection term (ECT) for 
cointegration with stock price is statistically significant at 5%. 
,6.LSP ,6.LOILP ,6.lTR ,6.LRGDP 
,6.LSP - 0.1486 -0.2197 0.0003 
[1.0910] [-0.3848] [-0.8062] 
,6.LOILP -0.0197 - -0.0269 0.0003 
[-0.3195 - [-0.0680] [-1.0863) 
,6.JTR -0.0121 -0.0078 - -2.99E-05 
[-0.8724] [-0.3680] - [-0.5466] 
,6.LRGDP 30.0907 4.3302 32.8301 -
[1.3721] [0.1296] [0.2342] -
ECT -0.0821 -0.0663 0.0659 0.0002 
[-3.6577]* [-1.9390]* [0.45992] [2.5847] 
Note: Number above the parentheses are coefficient values while numbers in the parentheses 
are t-statistic values. * indicates significance at 5%. 
The table shows that the enor correction term (ECT) for cointegration equation with stock price 
as dependent variable is statistically significant at 5%, meaning that oil prices have a long-run 
influence on stock prices. The ECT has a negative sign indicating a move back towards 
equilibrium and the coefficient estimate of the ECT of -0.082 shows that the relationship 
between stock prices and oil prices converges towards its long-run equilibrium at a slow speed. 
However, the short-run effect in the error correction model is tested using the Wald's test. The 
test works by testing the null hypothesis that a set of parameters is equal to the same value. In 
this model, the null hypothesis is that the short-run coefficients of oil price are simultaneously 
equal to zero. 
21 
Table 6: Wald Test result on short-run effects of oil prices on stock prices. From the table, 
the null hypothesis is rejected illustrating that oil prices have a significant impact on stock 
prices. 
Test Statistic Value Df Probability 
F -statistic 3.2425 (4, 122) 0.0145 
Chi-square 12.9699 4 0.0114 
Null Hypothesis: CC6)=C(7)=C(8)=C(9)=0 
Note: C (6), C (7), C (8) and C (9) are the short-run coefficients of oil prices. 
It was observed that the p-value was less than the general criterion of 0.05, so the null 
hypothesis is rejected indicating that in the short-run, oil prices have a significant influence on 
stock prices and this implies that there is a short-run causality running from oil prices to stock 
pnces. 
4.6 Impulse Response of Stock Prices due to Oil Price Shocks 
This section assesses the effects of oil price shocks on stock prices in terms of impulse response 
functions. IRFs illustrate the impact of a unit shock to the error of each equation of the V AR 
(Le & Chang, 2011 ), which is essentially mapping out the dynamic response path of a variable 
due to a one-period standard deviation to another variable. Information from application of 
IRFs should provide some further evidence on the patterns of linkages amongst stock prices 
and oil price shocks as well as contribute to enhancing our insights upon how other 
macroeconomic variables react to system wide shocks and how these responses propagate over 
time (Masih, 2011). Impulse response of stock prices is demonstrated by Table 7 while IRFs 
of the other variables in VAR system are provided for in Figure 2. 
22 
Table 7: Impulse Response of Stock Prices in Kenya due to oil price shocks over twelve 
month horizon. Table shows that stock prices are unresponsive to oil price shocks in the 
first month and begin responding from the second month and the response to oil price 
shocl{S gradually increases from the fourth month. 
Period LSP LOILP ITR LRGDP 
1 0.0577 0.0000 0.0000 0.0000 
2 0.0557 -0.0088 -0.0023 0.0065 
" 0.0514 -0.0021 0.0006 0.0124 .) 
4 0.0554 -0.0012 0.0006 0.01566 
5 0.0600 -0.0016 0.0065 0.0176 
6 0.0565 -0.0041 0.0056 0.0193 
7 0.0506 -0.0087 0.0080 0.0201 
8 0.0465 -0.0142 0.0092 0.0200 
9 0.0423 -0.0189 0.0124 0.0194 
10 0.0390 -0.0245 0.0130 0.0186 
11 0.0358 -0.0295 0.0144 0.0178 
12 0.0333 -0.0335 0.0154 0.0171 
Note: LSP (Log of stock prices); LOILP (log of oil prices); ITR (short-term interest rate); 
and LRGDP (log ofReal GDP). 
The results suggest that stock prices are not immediately responsive to innovations in oil prices 
in Kenya. The responses of stock prices to one standard deviation of oil price shock after one 
month is 0.0%. Six months after the oil price shocks, the responses of stock prices is -0.41% 
and after 12 months, the response of stock prices to oil price shocks is -3.35%. 
The impact of oil price shocks on stock prices increases over time and indicates a negative 
short-run effect of oil prices on stock prices. Additionally, the study shows that in the first 
month, short-term interest rates and real GDP do not affect stock prices. However after the 
second month, it can be seen that short-term interest rates and real GDP affect stock prices. 
The impact of real GDP on stock prices is continually positive in the first 12 months while the 
influence of short-term interest rates is sometimes negative and sometimes positive. 
4.7 Variance Decomposition of Stock Prices 
Due to its dynamic nature, variance decomposition accounts for the share of variations in the 
endogenous variables resulting from the endogenous variables and the transmission to all other 
variables in the system (Brooks, 20 14). In this study, the variance decomposition analysis 
specifically provides a tool for determining the relative importance of changes in oil prices, 
short-term interest rates and real GDP in explaining the volatility in stock prices. 
23 
Table 8: Variance decomposition of stock prices in Kenya over twelve month horizon. 
From the table it is seen that most of the variation in stock prices is due to its own shocks. 
Period LSP LOILP ITR LRGDP 
100.0000 0.0000 0.0000 0.0000 
2 98.0768 1.1922 0.0831 0.6479 
3 96.9652 0.8817 0.0627 2.0905 
4 95.8091 0.6631 0.0494 3.4784 
5 94.6588 0.5213 0.2956 4.5242 
6 93.5186 0.5114 0.3982 5.5718 
7 92.0445 0.7699 0.6195 6.5660 
8 90.2933 1.4544 0.8775 7.3749 
9 88.1322 2.5537 1.3252 7.9889 
10 85.5854 4.2461 1. 7506 8.4184 
11 82.6320 6.4568 2.2224 8.6888 
12 79.4565 9.0193 2.6961 8.8282 
Note: LSP (Log of stock prices); LOILP (log of oil prices); ITR (short-term interest rate); 
and LRGDP (log of Real GDP). 
Table 8 reveals that most of the variations in stock prices is due to the contribution of its own 
shocks. Immediately after the shock, oil prices contribute 1.19% to the variation in stock prices 
in Kenya. However, the effect decreases until the sixth month from where it begins to increase. 
After 12 months, the contribution of oil price shocks to the variation in stock prices is 9.02%. 
Moreover, the results show that real GDP explains stock price variation better than oil prices 
in the first 11 months while short-tem1 interest rates have the least effect on stock price 
variation. After 6 months, real GDP and short-term interest rates contribute 5.57% and 0.39% 
of the variation in the stock prices, respectively. 
24 
Chapter 5 Discussion 
5.1 Short-Run Effects of Oil Prices on Stock Prices 
The investigation on the sh01i-run effects of oil prices on stock prices is led by the estimation 
of a VECM which captures the short-run and long-run dynamic relationships of the variables. 
The study finds that in the short-run, oil prices have a significant influence on stock prices. 
Similarly, there is a short-run causality running from oil prices to stock prices. The result from 
IRF analysis indicates that oil price shocks causes an immediate decline in stock prices 
consistent with the finding ofPapapetrou (200 1) whose analysis revealed that an oil price shock 
has a negative effect on stock returns especially for the first four months. In addition, the 
variance decomposition of stock prices reveals that oil price shocks contribute 1.19% to the 
variation in stock prices in the short-run. On the other hand, in the short-run, the study finds 
that real GDP contributes 64.78% to the variation in stock prices, which is higher than the 
contribution by oil price shocks to the variation in stock prices. 
Similarly, Basher, Haug, & Sadorsky (2012) found that in the short-run, stock prices respond 
negatively to a positive oil price shock. This implies that in the short-run, oil price increases 
depresses stock prices, as was also observed by Filis, (20 1 0). 
5.2 Long-Run Effects of Oil Prices on Stock Prices 
The study seeks to find out the long-run effects of oil prices on stock prices in Kenya. Given 
the result of the unit root tests, the Johansen's cointegration test then reveals that the variables 
are cointegrated. The cointegration estimate shows that oil prices are a significant determinant 
of stock prices in the long-run, as was also observed by Miller & Ratti (2009); Narayan & 
Narayan (2010); Kapusuzoglu (2011); and Nwosa (2014). The VECM model reveals the 
existence of negative long-run effects of oil prices on stock prices in Kenya. From the variance 
decompostion analysis, the cumulative effects ofthe oil price shocks account for 9.02% of the 
variation in stock prices in the long-run. Similarly, Adaramola (20 12) observed a negative 
long-run relationship between oil prices and stock prices in Nigeria under his study. These 
results are consistent to theoretical expectations given the nature of Kenya as a net importer of 
oil. 
5.3 Conclusion and Recommendations 
The study examines the short-run and long-run effects of oil prices on stock prices in Kenya 
for the period spanning January 2003 to October 2015. Empirical evidence reveals that the 
short-run and long-run effects of oil prices on stock prices are statistically significant. Oil prices 
25 
have a negative influence on stock prices in Kenya in the short-run as indicated by the impulse 
response functions of stock prices. In addition, the reaction of stock prices to oil price shocks 
takes effect almost immediately. On the other hand, the results from the VECM show that there 
is a significant negative long-run relationship between oil prices and stock prices in Kenya. 
Given that the study finds the impact of oil price on stock prices to be significant, and Kenya 
being a net impmier of oil, the results of the paper are important in assessing the risk fi·om oil 
price changes and the impact oil price changes have on profits of firms in Kenya, which is 
therefore likely to play a significant role in the development of the country and its financial 
market. 
By the same token, the results in this paper are useful for individual and institutional investors, 
financial analysts, managers and policy makers who are concerned with oil price shocks in 
Kenya and thus should take into consideration changes in oil prices in their financial decisions 
since oil prices have significant effects on stock prices in Kenya. Furthermore, there is need 
for firms to adopt policies that utilize alternative sources of energy such as solar, wind and 
geothermal energy in order to reduce their dependency on oil for productive activities and limit 
the adverse effect of future oil price shocks. 
Fmiher research efforts could be channeled towards investigating on a sectoral perspective, the 
long-term effects of oil prices on stock prices from one industry to another and also provide 
opportunities for future research on the impact of oil price shocks and oil price risks on stock 
returns across various industries and in different countries especially Sub-Saharan Africa. 
Additionally, since oil prices have been found to have a significant impact on stock prices, it 
would be interesting from a practical perspective to consider the ability of oil price changes to 
forecast or predict stock price movements. Given that this study assummes linear linkages 
between the main variables, non-linear linkages should also be considered whose results may 
provide additional insight on the influence of oil prices on stock index returns. 
26 
References 
Adaramola, A. 0. (2012). Oil pnce shocks and stock market behaviour: the Nigerian 
experience. Journal of Economics, 3 (1), 19-24. 
Adebiyi, M. A. (2009). Oil price shocks, exchange rate and stock market behaviour: Empirical 
evidence from Nigeria. In Proceedings of the 14th Annual Cm?ference of the Aji'iccm 
Econometric Society. 
Al-Fayoumi, N. A. (2009). Oil prices and stock market returns in oil importing countries: the 
case of Turkey, Tunisia and Jordan .. European Journal of Economics, Finance and 
Administrative Sciences, 16, 86-101. 
Anoruo, E., & Mustafa, M. (2007). An empirical investigation into the relation of oil to stock 
markets prices. North American Journal of Finance and Banking Research, I (1), 22-
36. 
Basher, S. A., & Sadorsky, P. (2006). Oil price risk and emerging stock markets. Global 
Finance Journal, 17(2), 224-251. 
Basher, S. A., Haug, A. A., & Sadorsky, P. (2012). Oil prices,exchange rates and emerging 
stock markets. Energy Economics, 34(1), 227-240. 
Brooks, C. (2014)./ntroductmy econometricsfor.finance. Cambridge university press. 
Campbell, J. Y., Lo, A. W., & MacKinlay, A. C. (1997). The Econometrics of Financial 
Afarkets. Princeton: Princeton University Press. 
Carrasco, G. E. (2009). Selection of optimal lag length in co integrated V AR models with weak 
form of common cyclical features.lv!PRA, 1-24. 
Chang, Y., & Wong, J. (2003). Oil price fluctuations and Singapore economy. Energy Policy, 
31 (11), 1151-1165. 
Ciner, C. (2001). Energy shocks and financial markets: nonlinear linkages. Studies in Non-
Linear Dynamics and Econometrics, 5(3), 203-212. 
Cong, R. G. (2008). Relationships between oil price shocks and stock market:an empirical 
analysis from China. Energy Policy, 36(9), 3544-3553. 
Effiong, E. L. (2014). Oil price shocks and Nigeria's stock market: what have we learned from 
crude oil market shocks? OPEC Ener1::,ry Revie·w. 38(1), 36-58. 
27 
Enders, W. (2008). Applied econometric time series. John Wiley & Sons. 
Engle, R., & Granger, C. (1987). Co-integration and Error Correction: Representation, 
Estimation, and Testing. Econometrica, 55, 251-276. 
Fama, E. F. (1981). Stock returns, real activity, inflation, and money. The American Economic 
Review, 545-565. 
Filis, G. (20 1 0). Macro economy, stock market and oil prices: Do meaningful relationships 
exist among their cyclical t1uctuations? Energy Economics, 32(4), 877-886. 
Filis, G. D. (2011). Dynamic correlation between stock market and oil prices: The case of oil-
importing and oilexporting countries. International Revie·w of Financial Analysis. 
20(3), 152-164. 
French, K. R. (1987). Expected Stock Returns and Volatility . .Journal of Financial Economic. 
19(1), 3-29. 
Fliss, R. (2015, 11 10). Retrieved from http://www.empiwifo.uni-freiburg.de/lehre-teaching-
I/winter-term/ dateien-financial-data-analysis/ chapter4. pdf 
Gatuhi, S., & Macharia, P. (2013). Influence of Oil Prices on Stock Market Performance in 
Kenya. International .Journal ofk!anagment and Business Studies, 3(4), 59-65. 
Granger, C., & Weiss, A. (1983). Time Series Analysis of Error-Correcting Models. Studies in 
econometrics. time series, and multivariate statistics, 255-278. 
Hamilton, J.D. (1983). Oil and the macroeconomy since World War II. Journal o_fPolitical 
Economy, 228-248. 
Hamma, W. J. (2014). Effect of Oil Price Volatility on Tunisian Stock Market at Sector-level 
and Effectiveness of Hedging Strategy. Procedia Economics and Finance, 13, 109-127. 
Huang, R. D. (1996). Energy shocks and financial markets. Journal of Futures Markets. 16(1}, 
1-27. 
Johansen, S. (1988). Statistical analysis of cointegrating vectors. Journal o_f Economic 
Dynamics and Control. 12, 231-254. 
Johansen, S., & Juselius, K. (1990). Maximum likelihood estimation and inference on 
cointegration-with an application to the demand for money. Oxford Bulletin o.f 
Economics and Statistics. 52, 169-210. 
28 
Jones, C., & Kaul, G. (1996). Oil and the stock markets. 17w Journal ofFinance. 51 (2}, 463-
491. 
Kapusuzoglu, A. (2011). Relationships between Oil Price and Stock Market: An Empirical 
Analysis from Istanbul Stock Exchange (ISE). International Journal ofEconomics and 
Finance, 3(6), 99-106. 
Kwiatkowski, D.P. (1992). Testing the null hypothesis of stationarity against the alternative 
of a unit root, how sure are we that economic time series have a unit root? Journal of 
Econometric, 54, 159-178. 
Le, T. H., & Chang, Y. (2011). The Impact of Oil Price Fluctuations on Stock Markets in 
Developed and Emerging Economies. DEPOCEN, Working Paper Series No.23. 
MacKinnon, J. G. (1999). Numerical distribution .fimctions of likelihood ratio tests for 
cointegration (No. 9803). Department ofEconomics, University of Canterbury. 
Maghyereh, A. (2004). Oil price shocks and emerging stock markets: A generalized VAR 
approach. International Journal of Applied Econometrics and Quantitative Studies, 
1 (2), 27-40. 
Masih, R. P. (20 11 ). Oil price volatility and stock price fluctuations in an emerging market: 
Evidence from South Korea. Enert,ry Economics, 33(5}, 975-986. 
Miller, J. I., & Ratti, R. A. (2009). Crude oil and stock markets: Stability, instability, and 
bubbles. Energy Economics. 31 (4), 559-568. 
Misati, R.N. (2013). Commodity price shocks and inflation in a net oil-importing economy. 
OPEC Energy Review, 37(2), 125-148. 
Mork, K. A., Olsen, 0., & Mysen, H. T. (1994). Macroeconomic responses to oil pnce 
increases and decreases in seven OECD countries. The Energy Journal, 19-35. 
Nandha, M., & Hammoudeh, S. (2007). Systematic risk,and oil price and exchange rate 
sensitivities in Asia- Pacific stock markets. Research in International Business and 
Finance. 21,326-341. 
Narayan, P ., & Narayan, S. (20 1 0). Modelling the impact of oil prices on Vietnam's stock 
prices. Applied Energy, 87(1), 356-361. 
Nwosa, P. (2014). Oil prices and stock market price in Nigeria. OPEC Energy Revie·w, 38(1), 
59-74. 
29 
Oberndorfer, U. (2009). Energy prices,volatility and the stock market: Evidence from the 
Eurozone. Energy Policy, 37(12), 5787-5795. 
Ono, S. (20 11 ). Oil Price Shocks and Stock Markets in BRICs. The European Journal of 
Comparative Economics, 8(1), 29-45. 
Papapetrou, E. (200 1 ). Oil price shocks, stock market, economic activity and employment in 
Greece. Energy Economics. 23(5, 511-532. 
Parahoo, K. (2006). Nursing Research: Principles, Process and Issues. London, UK: Palgrave 
Macmillan. 
Park, J., & Ratti, R. (2008). Oil price shocks and Stock markets in the U.S. and 13 European 
Countries. Energy Economics, 30(5), 2587-2608. 
Pesaran, H. H., & Shin, Y. ( 1998). Generalized impulse response analysis in linear multivariate 
models. Economic letters 58(1), 17-29. 
Phillips, P ., & Perron, P. (1988). Testing for a Unit Root in Time Series regression. Biometrika, 
75, 335-346. 
Ross, S. A. ( 1976). The arbitrage theory of capital asset pricing .. Journal of economic theory, 
13(3), 341-360. 
Sadorsky, P. (1999). Oil price shocks and stock market activity. Energy Economics, 449-469. 
Vinh, X. (2014). An Empirical Investigation of Factors Affecting Stock Prices in Vietnam. 
Journal of Economics and Development. 16(1), 74-89. 
Zivot, E., & Wang, J. (2007). kfodelingfinancial time series vl'ith S-PiusCfi) (Vo/.191). Springer 
Science & Business Media. 
30 
Appendices 
Figure 2: Impulse Response Functions over twelve month horizon 
Response of LSP to LSP 
m~---------------. 
00~-----------------1 
·!11 
I 1 3 4 5 6 7 8 9 10 II 11 
Response of LOILP to LSP 
16~----------------, 
11 
m 
.04 
...._ __ _ 
I 1 3 4 5 6 7 8 9 10 II 11 
Response of ITR to LSP 
I 1 3 4 5 6 7 8 9 10 II 11 
Response oflRGOP to lSP 
I 1 3 4 5 6 7 8 9 10 11 12 
Response to Cholesky One S.D. Innovations 
Response oflSP to LOILP 
m~---------------. 
.00 
04 
I 1 3 4 5 6 7 8 9 10 II 12 
Response of lOilP to LOILP 
.16..---------------, 
.12~ 
re , "'----
00~-----------------4 
I 2 3 4 5 6 7 8 9 10 II 11 
Response ofiTR to LOILP 
----
1 2 3 4 5 6 7 8 9 10 II 12 
Response oflRGOP to LOILP 
.04..--------------, 
!11 
.01 
__...,... 
.oot---------""-----1 
I 2 3 4 5 6 7 8 9 10 II 12 
31 
Response of lSP to ITR 
.118.,.---------.., 
00 
.04 
·!11 
1 2 3 4 5 6 7 8 9 10 II 12 
Response of LOILP to ITR 
16..--------------, 
12 
re 
.04 
I 2 3 4 5 6 7 8 9 10 II 11 
Response of ITR to ITR 
I 2 3 4 5 6 7 8 9 10 II 12 
Response of LRGOP to ITR 
04..-------------, 
!11 
.01 
.OOt-------==:::---1 
--1 2 3 4 5 6 7 8 9 10 11 11 
Response of LSP to lRGOP 
re~--------------.., 
D6 
04 
-Jl2 
I 2 3 4 5 6 7 8 9 10 II 11 
Response oflOILP to lRGOP 
16..--------------, 
11 
re 
-00+--~~-----~ 
I 2 3 4 5 6 7 8 9 10 II 12 
Response of ITR to lRGOP 
1 1 3 4 5 6 7 8 9 10 II 12 
Response of lRGOP to LRGOP 
I 2 3 4 5 6 7 8 9 10 II 12