The relationship of world oil price and vietnamese oil companies stock return

Using multifactor model for the daily data during the period from 2010:1 to 2022:13 of Brent oil price changes, WTI crude oil price changes, Vietnamese oil companies stock price volatili

Dissertation submitted in partial fulfillment of the

Requirement for the MSc in Finance

FINANCE DISSERTATION ON THE RELATIONSHIP OF WORLD OIL

PRICE AND VIETNAMESE OIL

COMPANIES STOCK RETURN

NGUYEN THI THU HIEN

ID No: 22080926 Intake 6

Supervisor: Dr TRAN MANH HA

September 2023

EXECUTIVE SUMMARY

According to the data of Vietnamese State Stock Center, Vietnamese stock market capitalization was increasing sharply from 2,880,268 billion VND in 2016 to 9,309,889 billion VND in 2021 (over three times rise), compared to GDP in 2016 Vietnamese stock market capitalization to GDP was 68.7%, while in 2021 it was 148.84% (over 2 times increasing) Even though it was decreased in 2022 due to the Covid-19 pandemic and some other domestic macro economics events (such as bond booming and illiquidity of bond market together with the real estate booming), Vietnamese stock market is still expected to be attractive to the foreign investors when its P/E is 12 times, while Thai Land, Philippines, Malaysia or Indonesia is much higher than Vietnam at 16 times, ROE of VN-Index is 15% which is higher than other countries

in Asia at 9-10% (source: baochinhphu.vn)

The goal of the dissertation is to investigate the interactive relationships between oil price movement and Vietnamese oil companies stock returns Using multifactor model for the daily data during the period from 2010:1 to 2022:13 of Brent oil price changes, WTI crude oil price changes, Vietnamese oil companies stock price volatility, Vn-Index stock price risk, and exchange rate risk, it is found that only Brent oil price changes, Vn-Index stock volatility show statistically significant impact on the Vietnamese oil companies stock returns the WTI Crude oil price difference and exchange rate risk do not reveal the significant affect on the Vietnamese oil companies stock returns Increase in Brent oil price may increase the Vietnamese oil companies’ stock price and Vietnamese oil companies stock returns This outcome is in line with what is predicted by theory This suggests that petroleum stocks may be a good hedge against inflation

As oil prices increased, the Vietnamese stock market grew as well These results may be useful for individual and institutional investors, and policy makers

ACKNOWLEDGEMENTS

This thesis could not been completed without the support and encouragement of my mother and my passed away father, who always support me unconditionally in their own ways I would like to express my special thanks to my husband for his advocate to take care of my little family and my sweet children In addition, I would like to thank all my lecturers of the MSc joined program of Banking Academy of Vietnam and the University of the West of England for their extensive knowledge and skills Last but not least, I would like to express my sincere thanks to

my supervisor, Dr Tran Manh Ha for his valuable guidance, recommendation and time

TABLE OF CONTENTS

EXECUTIVE SUMMARY 2

ACKNOWLEDGEMENTS 3

TABLE OF CONTENTS 4

INTRODUCTION 5

LITERATURE REVIEW 9

METHODOLOGY 13

DATA 17

EMPIRICAL RESULTS 20

Histogram of IR and VNIR 20

Time series graphs for variables 20

Unit root test 21

Covariance and correlation 22

Parameter estimate, Durbin-Watson, AR, ARCH, GARCH, RAMSEY RESET 24

Homoskedasticity and heteroskedasticity 26

Prediction 27

CONCLUSION 31

REFERENCES 33

APPENDIX 36

INTRODUCTION

The world's economy is heavily dependent on crude oil, which is one of the most significant commodities in our economy It is nearly hard to find a factor that has a bigger impact on the global economy than oil Oil price variations cause economic recessions by reducing productivity, driving up inflation, and slowing down economic growth Costs for transportation and production go up as a result of rising oil prices and rising oil import expenses Assuming non-energy prices stay constant, this result in increased product prices, which in turn raise inflation Due to imbalances in the economy's supply and demand, increased oil prices and oil import costs cause GDP to decline This, in return, results in lower economic growth and recession Aside from that, the fluctuation in the price of oil raises risk and uncertainty, which has a negative influence on stock prices and decreases wealth and investment

For at least three reasons, the natural resource industry is a complex one First, it requires a lot of capital Building brand-new mines and pulp mills can cost billions of dollars Second, the resource base that natural resource corporations depend on is running out Natural resource extraction firms are constantly looking for low-cost natural resource deposits to exploit in order

to replace their dwindling asset base in order to stay in business Third, natural resource firms create a pretty uniform product, such as copper, gold, nickel, oil, and pulp

The energy sector has seen particularly difficult times during the past ten years Global petroleum demand and supply have been impacted by a number of political (the war in Ukraine) and economic (the Covid-19 epidemic) factors There is a direct effect brought on by Asia's decreased energy demand As major consumers of petroleum goods, the Asian nations' declining desire for these products lessened demand globally and subsequently lowered oil prices Large oil exporters like Russia and Venezuela experience budget deficits and decreased tax revenue as

a result of falling oil prices Investors fleeing risky assets move money into safe havens like US treasury bonds This will appreciate the US dollar, the denominated petroleum products, leads to consumers outside the US reducing their purchases (Blomberg & Harris, 1995), (Sadorsky, 2000)

Globalization has strengthened interdependencies across all economies in the globe by broadening the flow of products, services, and financial capital across national borders As a result of the rising importance of emerging economies like Brazil, China, and India, the expansion in global trade is more susceptible to increase in oil costs than in the past Due to the increased flow of portfolio capital (in the form of stocks, bonds, and mutual funds), both domestic and foreign investors are impacted by the effects of the oil price on emerging stock markets (Basher & Sadorsky, 2006)

Additionally, prior experience has demonstrated that the world's poorer nations are significantly more affected by oil price shocks Because the cost of importing oil skyrocketed due to the OPEC oil embargo of 1973, developing nations experienced significant economic and

social suffering The price of oil rose from $3 per barrel to $13 per barrel in only a few short months Developing nations needed loans from international lending institutions like the World Bank and the International Monetary Fund (IMF) in order to carry out their economic development initiatives (Rifkin, 2002) Commercial bank loans to emerging nations surged by 550% between 1973 and 1980 The second oil price shock of 1979 caused a global recession and made it harder for emerging nations to succeed since the cost of their oil imports increased while the cost of their other export goods decreased The third world's debt reached $1 trillion by 1985 The issue for the majority of developing nations was that any new borrowing was typically going toward paying off existing debt and purchasing imported oil There was hardly any money left over for brand-new economic development initiatives Today, there is a great deal of anxiety about the connection between high oil costs, high debt, and slow economic growth In the International Herald Tribune in 2000, Kofi A Annan, the Secretary General of the United Nations, stated that "debt-servicing costs are likely to increase if higher oil prices lead to higher international interest rates" in the ensuing years (Annan, 2000)

Oil is the key material in the modern economies Many industries use oil as their important and irreplaceable material, without oil, maybe many manufacturers stop processing and as a supply chain, the whole economy can be disrupted Especially for industrialization and modernization economy like Vietnam, oil consumption is growing fastest in the region, overtaking China, and rising by 7.5% annually over the last 20 years (according to statement of ANZ Bank on Vietnamnet on 2015) According to worldometer.info, as of 2016, Vietnam oil consumption was 478,000 barrels per day, ranking 34th of the oil consumption on the world, Vietnam oil production was 313,000 barrels per day, ranking 32nd in the world, meaning that domestic oil production was not enough for consumption, because of the high proven oil reserves

at 4.4 billion barrels in 2016, ranking 25th in the world, Vietnam was net exporter of oil However recently, oil production decreased from 17.23 million tons in 2016 to 10.97 million tons in 2021 (according to the report of vnexpress.net), the oil export decreased appropriately from 6.85 million tons in 2016 to 3.1 million tons in 2021, the oil import increased from 0.44 million tons in 2016 to 9.9 million tons in 2021 So the world oil price fluctuation will significantly affect the domestic market, especially stock market or stock return

The previous researches focused on the developed market such as the US and European countries (Park & Ratti, 2008), (Sadorsky, 2001), (Cong et al., 2008) The result is different Park & Ratti (2008) found that the relationship between oil price and oil-importing countries’ stock market is negative, while the relationship between oil price and oil-exporting countries’ stock market is positive Cong et al (2008) found that oil price shocks do not effect significantly

on the stock returns of most Chinese stock market indices However Narayan & Narayan (2010) found that oil prices had a positive and significantly impact on Vietnamese stock market

This research aims to analyze the relationship between Vietnamese oil companies’ stock return and the world oil price together with some variables such as: VN-Index stock returns, exchange rate from 2010:1 to 2022:13 in the situation that Vietnam is both oil-exporting and oil-importing country

Table 1 (in Appendix page 36) shows data on oil consumption for main areas and selected countries from 2010 to 2021 Asia Pacific showed the sharp increase in oil consumption with 25.54%, while Europe decreased significantly at 9.38% Compare country to country, Indonesia has decreased 3.86% while Vietnam experienced the strong increased at 29.61%

Table 2 (in Appendix page 39) shows data of oil production for main areas and selected countries from 2010 to 2021 While Asia Pacific leads the trend of oil consumption, it was near the bottom of oil production Vietnam increased significantly of oil consumption at 29.61% while decreased sharply in oil production at 39.24%

According to Vietnam Credit research on 2021, Oil and gas was the biggest industry from

2016 to 2020, contributing 10 percent to the country’s GDP

As forecasted by FiinGroup, the oil and gas industry profit after tax will grow by 741 percent

in 2021 The profit growth is predicted to be coming from the “downstream” group, including Binh Son Refining and Petrochemical Joint Stock Company, Vietnam National Petroleum Group, and Vietnam Oil Corporation Among these, Binh Son Refinery and Petrochemical have fulfilled 213 percent of the annual profit plan in the first quarter

Meanwhile, most enterprises in the “midstream” group have planned to reduce profits sharply in the context that oil and gas exploitation in Asia has not been accelerated, despite rising oil prices

In 2010, Vietnamese oil import was only USD0.5m, one sixteenth of export value But in

2020, oil import was nearly USD12m (increased sharply 24times) and nearly 3 times higher compare to the oil export value In 2021, the oil import and export value was decreased respectively but oil import value was still 3 times higher than the oil export value

According to Vietnam General Customs Office, in 2022, the percentage of oil export is 0.61% in total export value of Vietnam, the percentage of oil import is 2.16% in total import value of Vietnam as described in table 3 in the Appendix

The question is that Vietnam was on down trend in oil production while on up trend in oil consumption As a result, Vietnam has to import oil to meet the domestic demand And the world oil price surely has the impact on the Vietnam economy in general and Vietnam stock market in particular Oil price fluctuation can lead to the uncertainty in the input expenses such as transportation and logistics The stress from oil price affects the domestic consumer price index When the oil price increases, the Government normally will calm down other goods price in order to stable the inflation However in the emerging economy like Vietnam, when the oil price increases, the other commodity price will immediately surge, i.e electricity, labor cost or essential commodity prices, hence oil price movement will affect the whole economy

Brent oil is the light oil market in Europe, Africa, and the Middle East, exploited from oil fields in the North Sea between the Shetland Islands and Norway West Texas Intermediate (WTI) is the light oil market for US and sourced from US oil field in Texas Since Brent Crude oil is more widely used, it serves as the standard for the majority of oil prices—roughly two-thirds of all oil pricing (Administration for Energy Information) Because Brent Crude oil is produced close to the ocean, logistics expenses are considerably lower While WTI produced in landlocked regions, it leads to higher transportation expenses and more expensive price Because Vietnam is increasingly import oil, and the currency used is USD, the core currency of the world, the exchange rate of USDVND is also considered to have the impact on the Vietnam oil companies’ stock price change However the State Bank of Vietnam always control the USDVND exchange rate and keep it as stable as possible, the result may not reach the expected one

Increase in oil demand notwithstanding with the oil supply will lead to higher oil prices While oil is the blood of the economy, the higher the oil price will lead to more expensive input cost of enterprises and as consequences, higher selling price and inflation rate increases because

of lowering the amount of remaining income that consumers can use to purchase other goods and services and increasing the expenses of non-oil generating businesses and, decreasing profits and dividends, two major factors influencing stock prices Oil price volatility increases risk and uncertainty which negatively impacts stock prices and reduce investments

This thesis aims to find out the relationship of the world oil price change to Vietnam oil companies return by analyzing the stock market return, the world oil return, and the exchange rate change The methodology used is multifactor market model for several risk premiums The result is expected to that all the variables (the world oil price change, the market index return, and exchange rate) have the significant impact on Vietnam oil companies stock return This dissertation wants to contribute to the literature of the relationship between the world oil price risk, market index risk, exchange rate of USDVND change and the Vietnamese oil companies’ stock return

This thesis is organized as follows Section 2 reviews of the past literatures of the same topic Section 3 presents methodology and data used to estimate the model Section 4 reports the empirical results Section 5 concludes the topic

Sadorsky (2001)’s study used a multifactor market model in order to calculate the projected returns to stock prices in the Canadian oil and gas business He used monthly data of oil price, interest rate, market portfolio excess return and oil and gas share price return, all from period 1983-1999 Results are shown to demonstrate that stock price returns in the Canadian oil and gas business are significantly impacted by exchange rates, crude oil prices, and interest rates Particularly, a rise in the market or oil price element raises the return to Canadian oil and gas stock prices, whereas a rise in exchange rates or the term premium lowers the return Additionally, the oil and gas industry moves pro-cyclically and is less risky than the market Basher & Sadorsky (2006) examined the link between oil price risk and returns on emerging stock markets using an international multi-factor model that allows for both unconditional and conditional risk variables Thus, this study is one of the first in-depth ones to examine how oil price risk affects developing stock markets They used the international capital asset pricing model (CAPM), conditional and unconditional risk analysis and panel data of 21 emerging stock markets and the Morgan Stanley Capital International (MSCI) World Index In general, they found compelling evidence that the risk of rising oil prices affects the stock price returns in developing economies Other sources of unconditional risk like total risk, skewness and kurtosis have little impact on emerging market stock returns Results for conditional risk reveals some

interesting results For daily and monthly data, there is a positive and significant relationship between market betas and returns in up markets and negative and significant relationship between market betas and returns in down markets

On 2008, Park & Ratti researched about the oil price shocks and stock markets in the US and

13 European countries over 1986-2005 He used data of monthly data available over 1986–2005 for stock prices, short-term interest rates, consumer prices, and industrial production for the U.S and 13 European countries They used a vector autoregressive model (VAR) to capture the intricacies of the dynamic relationships between these variables and other factors, such as short-term interest rates, consumer prices, and industrial production, that may affect the relationships between oil price shocks and real stock returns As an oil exporter, Norway has a statistically significantly positive real stock return response to an increase in oil prices The median finding from variance decomposition analysis is that oil price shocks account for 6% of the volatility in real stock returns, which is statistically significant Increased oil price volatility severely reduces actual stock returns for many European nations but not for the United States More than interest rates, oil price shocks account for the variation in real stock returns in the United States and most other nations Within one or two months, an increase in the real price of oil is linked to a considerable rise in the short-term interest rate in the United States and eight out of thirteen European nations There is no evidence of asymmetric impacts on real stock returns of positive and negative oil price shocks for oil-importing European countries, in contrast to findings for the U.S and Norway

Cong et al (2008) investigates the relationship between oil price shocks and stock market in China by using multivariate vector auto-regression The majority of Chinese stock market indexes do not exhibit statistically significant effects of oil price shocks on real stock returns, with the exception of the manufacturing index and certain oil firms Some ''major'' oil price shocks cause a decline in oil firm stock values Increased oil volatility might lead to more speculative activity in the mining and petrochemicals indexes, which would improve stock returns Much more than interest rates for the manufacturing index can be explained by both the global and Chinese oil price shocks

Asteriou & Bashmakova (2013) assessed the impact of oil returns on emerging stock markets for ten Central and Eastern European Countries They looked at the connection between oil price risk and stock market returns for the developing capital markets of the Central and Eastern European Countries (CEECs) using a global multi-factor model For the time span from 22 October 1999 to 23 August 2007, a panel data technique is being used The fact that the oil price beta is statistically significant and negative suggests that the oil price does actually play a substantial role in predicting stock returns There is no statistically significant non-linear relationship between market risk and returns on emerging market stocks or between risk and returns on the oil price The analysis of conditional models reveals that emerging stock market returns respond favorably to increases in market returns When oil prices are low, the stock

returns' negative response to both upward and downward moves in the oil market is more pronounced

Diaz et al (2016) studied oil price volatility and stock returns in the G7 2014 (Canada, France, Germany, Italy, Japan, the UK, and the US) economies by utilizing monthly data from

1970 They take into account different parameters for oil prices (world, nominal, and real prices)

in order to quantify oil volatility With the following variables: interest rates, economic activity, stock returns, and volatility of the oil price, we estimate a vector autoregressive model while accounting for the structural break in 1986 They discover that the G7 stock markets react negatively to an increase in oil price volatility Additionally, the findings show that global oil price volatility has a greater overall impact on stock markets than domestic oil price volatility For the research of Vietnam stock market, Narayan & Narayan (2010) modeled the impact

of oil prices on Vietnam’s stock market Using daily data for the years 2000 to 2008, they also take into account the nominal exchange rate as an extra factor in determining stock prices They discover that nominal exchange rates, stock prices, and oil prices are all positively and statistically significant correlated with one another The outcome does not match what was predicted by theory As oil prices increased, the Vietnamese stock market grew as well Additionally, local market players shifted their preferences from holding foreign currencies and domestic bank deposits to stocks, and there was an increase in leveraged stock purchases as well

as investments made on behalf of relatives who were residing overseas It appears that these internal and domestic variables had a greater influence on the Vietnamese stock market than the increase in the price of oil

Nguyen & Bhatti (2012) modeled dependency between oil prices and stock markets in China and Vietnam by using nonparametric (chi- and K-plots) and parametric (copula) methods They noticed a left-tail relationship between global oil prices and the Vietnamese stock market, but the Chinese market exhibits the opposite behavior These findings give policymakers, investors, and risk managers dealing with these two markets a fresh understanding of how oil prices and stock markets behave, which has important ramifications

Some other researchers found that different industries were affected differently by oil price shocks in terms of stock prices A widely held opinion is that spikes in oil prices benefit upstream oil businesses, however has a negative impact on downstream companies and other industries Research by Huang et al (1996), for instance based on a VAR model and the correlative coefficient technique the S&P 500 index, 12 stock price indices for US industries, and three stock prices for oil companies Future returns on crude oil were discovered to adequately explain the stock returns of oil businesses, which may be viewed as their lead index, but had minimal impact on the overall market An expanded market model was utilized by Faff & Brailsford (1999) which they investigate various industry performance in Australian stocks market They discovered that the price of oil, the oil and gas business, and the various resources industry all had an impact on stock prices (Hammoudeh et al., 2004) using the Johansen co-

integration test discovered 1-month to 4-month WTI oil future price shocks explained oil extraction, refining, and marketing changes in a company's stock price VAR/VECM models were employed by Lanza et al (2005) to investigate the connections between six significant oil businesses, various stock exchanges, and the spread between crude oil future and spot prices He discovered that the oil firms' stock prices increased in direct proportion to the spread

In summary, the relationship of oil price movement and stock market in general or oil and gas industry stock returns or oil companies stock returns in particular is controversial and have different results in different countries or markets Whether this relationship presents in Vietnam stock market is discussed in this thesis This study estimates the impact of oil price risk on the Vietnam oil companies stock returns over 2010 to 2022

METHODOLOGY

This thesis aims to study the relationship of oil price changes and Vietnam Petroleum companies stock market returns In the situation that the decrease level in the oil production is much higher than the increase in the oil consumption (as shown in table 1 and 2, oil production

of Vietnam declines nearly 40% from 2010 to 2021 and oil consumption of Vietnam increase 29% for the same period) To secure the domestic oil demand, Vietnam is forced to import oil The Graph 5 shown that Vietnam oil import was triple the oil export in ton in 2021 So the author considers and tests the affect of the world oil price changes to the petroleum companies’ stock return, as well as the market return (VN-Index) on petroleum stock return

The previous researches focused on the developed market such as the US and European countries (Park & Ratti, 2008), (Sadorsky, 2001), (Cong et al., 2008) The result is different Park and Ratti (2008) found that the relationship between oil price and oil-importing countries’ stock market is negative, while the relationship between oil price and oil-exporting countries’ stock market is positive Cong et al (2008) found that oil price shocks do not effect significantly

on the stock returns of most Chinese stock market indices Sadorsky (2001) found the strong evidence that oil price risk impacts stock price returns in emerging market Narayan & Narayan (2010) found that oil prices had a positive and significantly impact on Vietnamese stock market The approach taken by using the multi-factor model

IRt = β0 + β1*VNIRt + β2*BRORt + εt (1)

In equation (1), IRt is the daily return of oil companies stock index, VNIRt is the daily return

on the market index, and BRORt is the daily return of Brent oil The parameters, β1 and β2 are the oil beta and the market beta, α is the intercept Equation (1) is estimated using OLS and model adequacy is checked using various regression diagnostic tests

Model (1) may be under/overestimated because the exchange rate factors are not included Exchange rate is added to model (1) for higher appropriate

IRt = β0 + β1*VNIR + β2*BROR + β3*WCOR + β4*EXR + εt (2) Model 2 is estimated by OLS and model adequacy is tested using regression diagnostic tests Recursive estimation techniques are used to assess the structural stability of the model

The steps to analyze the model and assess the relationship between Vietnamese oil companies’ return and VN-Index return, world oil return, as well as exchange rate are following Firstly, the author would like to run the histogram of Vietnam oil companies’ stock return and VN-Index stock return to find out whether it is normal distribution or not

Secondly, graphs for time series variables are shown to have a first sign of the time series data whether they are stationary or not and the range of deviation

Thirdly, the author will assess unit root test by Phillips – Peron (PP) root test and Augmented Dickey Fuller (ADF) root test1 to check whether a time series variable is non-stationary and possesses a unit root Depending on the test performed, the null hypothesis is typically the existence of a unit root, while the alternative hypothesis test either stationary, trend stationary or explosive root Most modeling strategies used in time series analysis focus primarily on the

data's stationary or unit root test The first step is to visually inspect and statistically verify the

series' attributes Graphs are the most basic tool for gaining a general understanding of the series' stationary However, statistical analyses are necessary for the ultimate judgment Unit root tests offer statistical proof of a series' stationary (Shrestha & Bhatta, 2018) The widely test used in stationary testing is Augmented Dickey Fuller (ADF) test and Phillips – Peron (PP) test The null hypothesis of ADF is random walk without drift If we do not reject null, the series is non-stationary whereas rejection means the series is stationary (Dickey & Fuller, 1979)

The main difference between the Augmented Dickey – Fuller and Phillip – Peron tests is that

PP is a non-parametric test, therefore, under the null hypothesis, it is not necessary to characterize the shape of the serial correlation of ∆yt As a result, the method for calculating the t-ratio to determine the value of p is altered Additionally, PP adjusts the statistics to take into account the problems with autocorrelation and heteroskedasticity The method of assessing the hypothesis is comparable to the ADF test

Fourthly, the covariance2 and correlation, which is very important in the model, is tested whether each variable depend on each other and related to each other Covariance of the two variables reflects the linear correlation of two variables Covariance can be positive, negative or

0, reflecting a positive or negative relationship between two variables The smaller the absolute value of the covariance is, the lower the strength of the relationship is The covariance between two variables is the determining factor of the moment correlation coefficient that holds them (Hill et al., 2017) If the covariance between the two variables is positive, then when the x value

is above its mean, the y value also tends to be above its mean, and when the x value is below its mean, the y value also tends to be less than its mean In this instance, it is claimed that the random variables X and Y are directly or positively related The connection is inverse or negative if σXY < 0 There is neither a positive nor a negative relationship if σXY = 0

Because it is unit free and falls within the range [-1, 1], correlation is the chosen measure when there is some degree of linear association, whereas the size of a covariance will depend on

the units of measurement of the two variables Correlation matrix is to check the existence and

strength of the relationship between the variables

1 The formula of ADF and PP test are shown in Appendix page 44

2 The formula of Covariance is shown in Appendix page 44

For two random variables X, Y, their correlation is calculated as below:

model's autoregressive (AR) nature implies that each variable's present value is governed by its

previous value plus a few adjusting parameters These adjustment factors are calculated using the relationship between the current value and previous values The important testing is to test the stationary of the variables If a time series data's value tends to return to its long-run average value and its characteristics are not solely impacted by the passage of time, it is said to be stationary The non-stationary time series, on the other hand, has a tendency to deviate from its long-run average value; as a result, its mean, variance, and covariance change over time (Hamilton, 1994)

The random error at time t is connected to the random error in the preceding time period plus

a random component, according to this model There isn't an intercept parameter, in contrast to the AR model, because xt has a mean of zero (Hill et al., 2017)

We continue to test ARCH (Autoregressive Conditional Heteroskedasticity) Autoregressive time series approach to modeling volatility that changes over time (conditional heteroscedasticity) ARCH model will test for time dependency and “clustering” of volatility Because it makes sense to describe volatility as a function of errors et, the ARCH model has an intuitive appeal Financial experts refer to these mistakes as "news" or "shocks" rather frequently They stand for the unanticipated! The ARCH model states that series volatility increases with shock magnitude This model also accounts for volatility clustering since large changes in et are fed into larger changes in ht via the lagged effect of et-1 ARCH(1) and ARCH(4) are lagrange multiplier test statistics for residual ARCH effects at lags 1 and 4 (Engle, 1982) Autoregressive Conditional Heteroskedasticity at lag 1 can be modeled as below

3 The formula of Durbin-Watson is presented in Appendix page 44

The null hypothesis H0: no ARCH effect/ H1: ARCH (p) disturbance

GARCH (p, q) model (Generalized Autoregressive Conditional Heteroskedasticity) is an improved method of ARCH model when ARCH model is difficult to predict at high order The GARCH (p, q) model can be write as

Ht = δ + β1e2

t-1 + β2ht-1

Where p is the total number of lagged h terms and q is the total number of lagged e2 terms

We also observe that in order to achieve stationary, β1 + β2 < 1 is required; otherwise, we have an

"integrated GARCH" process, or IGARCH Due to how well it fits a variety of data series, the GARCH (1, 1) model is a particularly popular specification It reveals that the volatility fluctuates in response to lag shocks (e2

t-1), but that the system also has momentum operating via

ht-1 Moreover, this model's ability to capture extended lags in the shocks with a minimal number

of parameters is one factor in its popularity An ARCH (q) model requiring the estimate of (q + 1) parameters, where q is large, say q ≥ 6, can capture equivalent effects as a GARCH(1, 1) model with three parameters (δ, β1, and β2) (Lundbergh & Teräsvirta, 2002) (Hill et al., 2017) RAMSEY RESET4 test for omitted variables, developed by (Ramsey, 1969) RESET (Regression Specification Error Test) is designed to detect omitted variables and incorrect functional form or model Testing for model misspecification is one method of determining whether our model is sufficient or whether it needs improvement It might be incorrectly described if significant variables were left out, unimportant ones were included, the wrong functional form was used, or the multiple regressive model underlying assumptions were broken

by the model

After these tests were implemented, if there is no amendment needed, the model was analyzed to study the relationship of Vietnamese oil companies’ stock return, VN-Index return, Brent oil return, WTI oil return and exchange rate change

4 The formula of RAMSEY RESET test is presented in Appendix page 45

DATA

Data is time series data in daily and covers the period from 2010:1-2022:13 (3,194

observations), all are from Investing.com The purpose of this thesis is to analyze the change of oil companies’ stock price, so the dependent variable is the return of the Vietnam oil companies’ stock price from 2010 to 2022 The data is taken from 2010, after the financial crisis 2008-2009, and 2010 was the year of the world economy recovered with more stable trend The Vietnam oil companies’ stock return is calculated based on the daily closed price of Petroleum industry stock price as follow:

r = Rf + β*(Rm – Rf)

Where,

r is the expected return of the stock

β is the beta of the stock, measures the volatility of stock returns to the market risk

Rf is the risk free rate

Rm is the expected return of the market

Basically, the stock return totally depends on the market return The first and important independent variable is market return, which is taken by the change of today closed price of VN-Index and the previous day closed price of VN-Index

Because the oil companies stock return is being evaluated, and with the upward trend of more import than export oil, Vietnam oil depends on the world oil The world oil price including WTI crude oil and Brent Oil price change are also considered to have the impact on the Vietnam oil companies’ stock price change The changes in Brent/ WTI Oil price based on the daily closed price of future Brent/ WTI Crude Oil taken from Bloomberg.com Instead of using spot prices, which are more susceptible to short-term price swings brought on by blips in supply or demand, one instead utilizes future pricing While WTI is the benchmark for the US light oil market, Brent crude is utilized for the worldwide light oil market, which includes Europe, Africa, and the Middle East (Sadorsky, 2001) analyzed the relationship between the stock returns of Canadian

oil and gas companies and the risk factors, including the oil prices changes (Basher & Sadorsky, 2006) studied the oil price risk and emerging stock markets (Asteriou & Bashmakova, 2013) evaluated the impact of oil returns on emerging stock markets (Eastern European countries) (Cong et al., 2008) assessed the relationships between oil price shocks and stock market, the case

of China (Narayan & Narayan, 2010) modeled the impact of oil prices on Vietnam’s stock prices

The oil import and export of Vietnam will be impacted by the change of exchange rate This

is exchange rate USDVND which is most relevant to Vietnam companies because USD is the denominated currency all over the world Foreign exchange risk is modeled using the exchange rate variable For global corporations, exchange rate risk may be particularly significant, especially at companies that deal with natural resources For instance, (Louden, 1993) and (Khoo, 1994) both discovered a correlation between changes in the Australian gold stock and changes in the currency rate Overall, little research has been done on how foreign exchange risk exposure affects oil and gas enterprises The exchange rate change based on the daily exchange rate of USDVND end of day, as quoted by the State Bank of Viet Nam

With data collected, the summary statistics for the data are presented in Table 1 below Variale Mean Std Dev t-statistic Kurtosis Skewness

IRt 0.0003429 0.0199052 4.76682 -0.1481452 VNIRt 0.0002817 0.0120212 60.25 6.131682 -0.6706471 BRORt 0.0003052 0.0238801 2.64

WCORt -0.0008002 0.0648165 0.29

EXRt 0.000106 0.0067125 1.42

Table 1: Summary statistics of data

The average return of Vietnam oil companies’ return is higher than the Vn-Index with 0.034% and 0.028% accordingly With the higher return, the standard deviation of Petroleum companies is higher than the standard deviation of VN – Index with 2% and 1.2% respectively, meaning that the Petroleum companies’ stock price had a wider range of price than VN-Index stock price or the Petroleum companies stock price is riskier of volatility than VN-Index stock price The standard deviation5 of return of WCO (WTI Crude Oil) is the highest number with 6.48%, meaning that the WCO price is the variable with highest volatility of change The return

of BRO (Brent Oil) stays at the second highest standard deviation of risk with 2.39% The exchange rate variance changed slightly during the period with 0.67% volatility The bigger the deviation within the data collection, the further the data points deviate from the mean; hence, the higher the standard deviation, the more dispersed the data

5 The definition and formula of standard deviation is presented in Appendix page 42

T-statistic6 of VN-Index return and Brent Oil return does have a significant mean, while WTI Crude Oil and Exchange rate difference does not have a significant mean This can rise very first awareness that VN-Index return and Brent Oil return will significantly impact on the Vietnamese oil companies’ stock return

IR’s skewness7 is -0.1481452, VNIR’s skewness is -0.6706471, they are skew on the left; the data is sparse on the left and piled up at the right end of the distribution IR’s kurtosis is 4.76682, VNIR’s kurtosis = 6.131682, they are higher than 3, it is said to be leptokurtosis or fatter tail The kurtosis of stock market return is higher than the kurtosis of Vietnam oil companies return, meaning that the stock market return has the higher risk than the Vietnam oil companies return Base on the histogram of IR and VNIR, both of the two variables are normal distribution

6 The definition and formula are presented in Appendix on page 43

7 The definition and formula of skewness and kurtosis are presented in Appendix on page 43

EMPIRICAL RESULTS

Histogram of IR and VNIR

Graph 1: Histogram of IR and VNIR

The IR’s histogram seems normal distributed with evenly data around zero value While the VNIR’s histogram is more on the left with negative value, and the kurtosis of VNIR is higher than the IR’s kurtosis meaning that the return of VNIR is more risky but investor have more chance to get better return

Time series graphs for variables

Graph 2: Times series graph for variables

Data of IR fluctuates more than the VNIR with the highest points are in 2012, 2016, 2018 and 2022 at around 6% to 7% The lowest points are in 2018, 2020, 2021 and 2022 with value around -7% to -8% This creates the opportunity for investors to get higher returns than the VN-Index, but they can bear larger loss than the VN-Index For the data of VNIR, its range from -6%

to 5%, the highest points are in 2020 and 2022 at 5%, the lowest points are in 2011 and 2021 at around -6% The data of BROR oscillates quite small but significantly high and falls in 2020 with the highest mark is 4%, the lowest mark is -2.1% This is because the Covid-19 has just started, the investors’ fear of falling price in future oil price, they cut lost and other investors bought in when the OPEC stated that they will increase the quantity of the oil The data of WCOR is quite stable but slightly rise at 0.2% and sharply drop in 2020 at -3% This happens only one time during 13 years from 2010 to 2022 The data of EXR fluctuates strongly during the period but quite stable in 2022 The highest point is in 2011 at 6.2%, the second highest point is

in 2016 at 4% The lowest point is in 2016 at -3.8%

Unit root test

Table 2: Unit root tests

*** denotes the statistic is significant at 1% confidence level All the P value of ADF root test and PP root test is 0 < 0.5, so the null hypothesis “Random walk without drift” is rejected, meaning that all the variables are stationary

These results match with the graph of time series of data as the data is fluctuate around 0 This means that the Vietnamese oil companies’ stock price, VN-Index stock price and Oil price are sometimes increased and sometimes decreased Hence, the return of Vietnamese oil companies’ stock, VN-Index stock and Oil are sometimes positive and sometimes negative

Covariance and correlation

Table 3: Covariance matrix

Cov (VNIR, BROR) and Cov (VNIR, WCOR) are negative, meaning that when VNIR is above the expected return, the BROR/ WCOR are below the expected return, or it can be understood that when Brent Oil/ WTI Oil price decrease, the return of VN-Index is expected to increase It may be helpful for investors when they see the sign of the world oil price falling, they may expect the stock will rise and they may buy in

Cov (VNIR, EXR) is positive, meaning that when EXR is higher than the expected return, the VNIR is higher than the expected return, or when the exchange rate of VND/USD increase or USD is appreciated, the VNIR is expected to increase as well