Mathematics Analysis and Approach Standard Level Internal Assessment:
Multiple Regression Analysis between Crude Oil Price and Geopolitics
Introduction
7 years ago, my native country Ukraine engaged in war. Before we used to keep away from
politics, however when geopolitical events affect you personally, it is hard to be apolitical.
Since then, I have developed a big interest in international relations, especially conflicts and
economics, and hope to pursue it in my future. Throughout IB I explored this area, and once
stumbled upon an article about how Geopolitical Tensions influence Oil Prices. While
reading the article, it was hard for me to comprehend that such things as terrorism can
influence Oil Prices. Therefore, I started to wonder What is the level of relationship between
Geopolitical Tension and Crude Oil Price, and how can we estimate future oil prices’ shifts
depending on geopolitical factors? Since Oil is the most important source of energy, it is
crucial to know what factors affect its price change. I want to know how the sphere of my
interest – geopolitics – can influence the economy of Oil. The use of mathematics through
Multiple Linear Regression plays a pivotal role in understanding the relationship and project
possible oil changes.
Background information
Crude Oil Prices and influential factors
Crude Oil has been the most essential commodity in the world during the last half-century, as
the majority of our energy comes from the oil supply. Its production, storage, import, and
export create a multi-trillion-dollar industry. However, the prices of Crude Oil are a
complicated puzzle that is influenced by multiple factors. In this exploration, I am focusing
on geopolitical factors - Number of Terrorist Attacks Worldwide and OPEC’s Oil Production
,in Barrels. OPEC is a major oil-related intergovernmental organization, as its members own
the biggest amount of oil (in barrels) in the world.1
Historically, the relationship between geopolitics and fuel prices is a sensitive one and, it is
based on principles of supply and demand. As some of the biggest oil-producing countries
suffer from high political instability, the oil market is threatened by the rising number of
terrorist attacks that change physical supply or the sheer threat of production instability which
decreases demand. These geopolitical metrics thus allow capturing hidden structure in the
dynamics of the Crude Oil market.2
Econometrics
To understand the relationship between the oil market and geopolitical changes and to
explain consistently occurring correlation, econometrics is used. In general, the application of
econometrics serves to develop theories or test existing hypotheses and forecast future trends
from historical data in economics. It is useful for the investigation as it subjects real-world
data, as well as provides models to analyze the complex relationship between many variables.
This exploration will rely on multiple regression analysis techniques and null-hypothesis
testing. 3
Multiple Linear Regression
In statistics, linear regression is a model of a relationship between one independent variable
and a dependent variable. However, in reality, multiple factors predict the outcome of an
1
Mostert, Wolfgang. “Factors Influencing the International Price of Oil in the Medium to Long-Term.”
Mostert.dk. Accessed January 15, 2021. https://mostert.dk/pdf/Factors%20influencing%20price%20of
%20oil.pdf.
Niemi, Jaradd. “T-Distribution Critical Value.” Accessed March 19, 2021.
https://www.jarad.me/courses/stat226/tables/t-table.pdf.
2
Bariviera, Aurelio. “Crude Oil Market and Geopolitical Events: an Analysis Based on Information-Theory-
Based Quantifiers.” Https://Arxiv.org/. Accessed January 5, 2021. https://arxiv.org/pdf/1704.04442.pdf.
3
Hayes, Adam. “Econometrics: What It Means, and How It's Used.” Investopedia. Investopedia,
December 14, 2020. https://www.investopedia.com/terms/e/econometrics.asp.
,event. A dependent variable is rarely explained by only one independent variable. The crude
oil price depends on more than just the amount of oil produced or just the demand for it. To
assess the strength of a relationship, in which more than two independent variables are
present, which is often the case in social science, multiple linear regression is used. In causal
analysis, it also separates the effects of independent variables on the dependent variable so
that you can examine the unique contribution of each variable. The model assumes that there
are no major correlations between the independent variables, and often the effect of other
predictors statistically eliminated.
The formula for predicting the value of a dependent variable by multiple independent
variables:4
y=β 0 + β 1∗x 1 + β 2∗x 2+ … β p∗x p+ ϵ where,
y=dependent variable
x=independent variables
β 0= y −intercept (constant term)
β p=slope coefficients for each ∈dependent variable
ϵ=themodel ’ s error t erm(residuals)
Aim of Exploration
The study aims to “Detect whether two independent variables (Number of Terrorist Attacks
Worldwide and OPEC’ Oil Production in barrels) affect the dependent variable (Price of
Crude Oil), to estimate the magnitude of this effect, if any, and to create forecast formula
through Multiple Regression Analysis”.
Results Expected
4
Bevans, Rebecca. “Multiple Linear Regression: A Quick and Simple Guide.” Scribbr, October 26, 2020.
https://www.scribbr.com/statistics/multiple-linear-regression/.
, Moderate correlation between the independent variables and the dependent variable is
expected, as the variables are produced by human-behavior. In the exploration, independent
variables are Number of Terrorist Attacks ( x 1 ¿ and OPEC’s Oil Production in barrels ( x 2 ¿ ,
the dependent variable is Price of Crude oil ( y ) .
Alternative hypothesis (H A ) anticipated to be accepted.
H 0 :there is no relationship between the independent variables∧thedependent variable ( y ) ,
¿ the independent variables should not remain∈themodel .
H A :there is a significant relationship between theindependent variables ( x 1 , x 2 )∧the
dependent variable ( y ) ,∧theindependent variables should remain∈the model .
Raw Data
To conduct the multiple regression analysis, quantitative data for the two independent
variables ¿ ¿) number of terrorist attacks worldwide5 and number of barrels of crude oil
produced by OPEC between 1989-20196, and data for dependent variable ( y ¿Price of Crude
between 1989-20197 was found. In total, every variable includes 30 independently selected
observations that can be observed in Table 1. First 10 observations are presented below.
Table 1. Data on number of terrorist attacks worldwide ( x 1 ¿ ; OPEC’s production of oil barrels ( x ¿ ¿ 2 and Price of Crude
Oil ( y ) between 1989-2019
Raw Data (sample)
Year Number of Terrorist OPEC’s Crude Oil Price of Crude Oil addjusted
5
Hayes, Adam. “How Multiple Linear Regression Works.” Investopedia. Investopedia, March 17, 2021.
https://www.investopedia.com/terms/m/mlr.asp.
6
Global Terrorism Database. “Terrorism. Number of Terrorist Attacks, 1970-2019.” Accessed March 13,
2021. https://ourworldindata.org/terrorism.
7
Sönichsen, N. “OPEC's Crude Oil Production in Barrels from 1989 to 2019.” Accessed March 10, 2020.
https://www.statista.com/statistics/292590/global-crude-oil-production-opec-share/.
Multiple Regression Analysis between Crude Oil Price and Geopolitics
Introduction
7 years ago, my native country Ukraine engaged in war. Before we used to keep away from
politics, however when geopolitical events affect you personally, it is hard to be apolitical.
Since then, I have developed a big interest in international relations, especially conflicts and
economics, and hope to pursue it in my future. Throughout IB I explored this area, and once
stumbled upon an article about how Geopolitical Tensions influence Oil Prices. While
reading the article, it was hard for me to comprehend that such things as terrorism can
influence Oil Prices. Therefore, I started to wonder What is the level of relationship between
Geopolitical Tension and Crude Oil Price, and how can we estimate future oil prices’ shifts
depending on geopolitical factors? Since Oil is the most important source of energy, it is
crucial to know what factors affect its price change. I want to know how the sphere of my
interest – geopolitics – can influence the economy of Oil. The use of mathematics through
Multiple Linear Regression plays a pivotal role in understanding the relationship and project
possible oil changes.
Background information
Crude Oil Prices and influential factors
Crude Oil has been the most essential commodity in the world during the last half-century, as
the majority of our energy comes from the oil supply. Its production, storage, import, and
export create a multi-trillion-dollar industry. However, the prices of Crude Oil are a
complicated puzzle that is influenced by multiple factors. In this exploration, I am focusing
on geopolitical factors - Number of Terrorist Attacks Worldwide and OPEC’s Oil Production
,in Barrels. OPEC is a major oil-related intergovernmental organization, as its members own
the biggest amount of oil (in barrels) in the world.1
Historically, the relationship between geopolitics and fuel prices is a sensitive one and, it is
based on principles of supply and demand. As some of the biggest oil-producing countries
suffer from high political instability, the oil market is threatened by the rising number of
terrorist attacks that change physical supply or the sheer threat of production instability which
decreases demand. These geopolitical metrics thus allow capturing hidden structure in the
dynamics of the Crude Oil market.2
Econometrics
To understand the relationship between the oil market and geopolitical changes and to
explain consistently occurring correlation, econometrics is used. In general, the application of
econometrics serves to develop theories or test existing hypotheses and forecast future trends
from historical data in economics. It is useful for the investigation as it subjects real-world
data, as well as provides models to analyze the complex relationship between many variables.
This exploration will rely on multiple regression analysis techniques and null-hypothesis
testing. 3
Multiple Linear Regression
In statistics, linear regression is a model of a relationship between one independent variable
and a dependent variable. However, in reality, multiple factors predict the outcome of an
1
Mostert, Wolfgang. “Factors Influencing the International Price of Oil in the Medium to Long-Term.”
Mostert.dk. Accessed January 15, 2021. https://mostert.dk/pdf/Factors%20influencing%20price%20of
%20oil.pdf.
Niemi, Jaradd. “T-Distribution Critical Value.” Accessed March 19, 2021.
https://www.jarad.me/courses/stat226/tables/t-table.pdf.
2
Bariviera, Aurelio. “Crude Oil Market and Geopolitical Events: an Analysis Based on Information-Theory-
Based Quantifiers.” Https://Arxiv.org/. Accessed January 5, 2021. https://arxiv.org/pdf/1704.04442.pdf.
3
Hayes, Adam. “Econometrics: What It Means, and How It's Used.” Investopedia. Investopedia,
December 14, 2020. https://www.investopedia.com/terms/e/econometrics.asp.
,event. A dependent variable is rarely explained by only one independent variable. The crude
oil price depends on more than just the amount of oil produced or just the demand for it. To
assess the strength of a relationship, in which more than two independent variables are
present, which is often the case in social science, multiple linear regression is used. In causal
analysis, it also separates the effects of independent variables on the dependent variable so
that you can examine the unique contribution of each variable. The model assumes that there
are no major correlations between the independent variables, and often the effect of other
predictors statistically eliminated.
The formula for predicting the value of a dependent variable by multiple independent
variables:4
y=β 0 + β 1∗x 1 + β 2∗x 2+ … β p∗x p+ ϵ where,
y=dependent variable
x=independent variables
β 0= y −intercept (constant term)
β p=slope coefficients for each ∈dependent variable
ϵ=themodel ’ s error t erm(residuals)
Aim of Exploration
The study aims to “Detect whether two independent variables (Number of Terrorist Attacks
Worldwide and OPEC’ Oil Production in barrels) affect the dependent variable (Price of
Crude Oil), to estimate the magnitude of this effect, if any, and to create forecast formula
through Multiple Regression Analysis”.
Results Expected
4
Bevans, Rebecca. “Multiple Linear Regression: A Quick and Simple Guide.” Scribbr, October 26, 2020.
https://www.scribbr.com/statistics/multiple-linear-regression/.
, Moderate correlation between the independent variables and the dependent variable is
expected, as the variables are produced by human-behavior. In the exploration, independent
variables are Number of Terrorist Attacks ( x 1 ¿ and OPEC’s Oil Production in barrels ( x 2 ¿ ,
the dependent variable is Price of Crude oil ( y ) .
Alternative hypothesis (H A ) anticipated to be accepted.
H 0 :there is no relationship between the independent variables∧thedependent variable ( y ) ,
¿ the independent variables should not remain∈themodel .
H A :there is a significant relationship between theindependent variables ( x 1 , x 2 )∧the
dependent variable ( y ) ,∧theindependent variables should remain∈the model .
Raw Data
To conduct the multiple regression analysis, quantitative data for the two independent
variables ¿ ¿) number of terrorist attacks worldwide5 and number of barrels of crude oil
produced by OPEC between 1989-20196, and data for dependent variable ( y ¿Price of Crude
between 1989-20197 was found. In total, every variable includes 30 independently selected
observations that can be observed in Table 1. First 10 observations are presented below.
Table 1. Data on number of terrorist attacks worldwide ( x 1 ¿ ; OPEC’s production of oil barrels ( x ¿ ¿ 2 and Price of Crude
Oil ( y ) between 1989-2019
Raw Data (sample)
Year Number of Terrorist OPEC’s Crude Oil Price of Crude Oil addjusted
5
Hayes, Adam. “How Multiple Linear Regression Works.” Investopedia. Investopedia, March 17, 2021.
https://www.investopedia.com/terms/m/mlr.asp.
6
Global Terrorism Database. “Terrorism. Number of Terrorist Attacks, 1970-2019.” Accessed March 13,
2021. https://ourworldindata.org/terrorism.
7
Sönichsen, N. “OPEC's Crude Oil Production in Barrels from 1989 to 2019.” Accessed March 10, 2020.
https://www.statista.com/statistics/292590/global-crude-oil-production-opec-share/.
