THE IMPACT OF MACROECONOMIC VARIABLES ON THE STOCK MARKET OF

THE OIL-EXPORTING COUNTRIES: THE CASE OF IRAN

Dissertation presented to the

School of Management

Alliant International University

San Francisco

In partial fulfillment of the requirements for the degree of

Doctor of Business Administration

by

Naser Niknam Esfahani

San Diego

2016

Approved by:

Dr. Mustafa Sayim

Dr. Louise Kelly

Dr. Yousef Ibrahim

 ProQuest Number: 10243507

All rights reserved

INFORMATION TO ALL USERS

The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.

ProQuest 10243507

Published by ProQuest LLC ( 2016 ). Copyright of the Dissertation is held by the Author.

All rights reserved.

This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.

ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
MACROECONOMICS AND THE STOCK MARKET ii

© Naser Niknam Esfahani, 2016

MACROECONOMICS AND THE STOCK MARKET iii

Dedication

I dedicate this dissertation to my lovely wife Marzi, my son Ahoura, and my daughter

Chista. This achievement could not have been accomplished without their support.

Also to my parents who have always supported me unconditionally.

MACROECONOMICS AND THE STOCK MARKET iv

Acknowledgments

I would like to express my appreciation to my committee for continued support and

encouragement. My special thanks goes to my committee chairperson Dr. Mustafa Sayim for his

patience, motivation, and immense knowledge. His advice helped me all the time during the

research and writing this thesis. Also, I would like to thank Dr. Louise Kelly, and Dr. Yousef

Ibrahim, the members of my committee for taking part in this research and for their insightful

comments and encouragement.

MACROECONOMICS AND THE STOCK MARKET v

Abstract

The purpose of this study is to investigate the impact of oil income and the gold price fluctuation

on the stock market of Iran, and to examine whether consideration of the oil and the gold markets

followed by other investment markets, such as the real estate market and the foreign exchange

market have any significant impact on the stock market behavior. This study examines the

impact of the oil income and gold price fluctuation, as well as other macroeconomic indicators

such as housing construction, deposit interest rate, foreign exchange rate, consumer price index,

and gross domestic products on the Iranian stock market, as measured by the quarterly data from

January 1990 to December 2012. First, regression analysis was applied for oil revenue, housing

activities, deposit interest rate, foreign exchange rate, CPI, and GDP to examine the stock market

reaction to each market activity. After this, ceteris paribus, oil revenue was replaced with gold

price to examine whether the price of gold without consideration of the oil revenue had any

impact on the stock market activity. Finally, the model was run to examine the impact of all

variables, including oil revenue and gold price, upon stock market activity. In a separate model,

this paper also examines the relationship between all of the above-mentioned variables and stock

market volatility. In the next phase, the impulse response function based on the vector auto-

regression model was generated to examine the behavior of the stock market followed by a shock

to each independent variable. Because Iran is an oil-exporting country—not an oil-importing

country—economic forces, market structure as well as the investment culture and investing

policies are different in Iran than in Western nations or other oil-importing countries, it may not

be wise or productive to analyze investment in the Iranian stock market from the same

perspective or with the same methods used when analyzing investment in Western countries.

Because gold, foreign currencies, and the real estate market are the most popular and common
MACROECONOMICS AND THE STOCK MARKET vi

investment markets in Iran for domestic investors, this paper aims to determine whether there is

any relationship between those markets and the stock market in general from an investment point

of view.
MACROECONOMICS AND THE STOCK MARKET vii

Table of Contents

Dedication iii

Acknowledgments iv

Abstract v

List of Tables xi

List of Figures xiii

CHAPTER I. Introduction 1

Background of the Problem 3

Purpose of the Study 5

Importance of the Study 6

Scope of the Study 7

Research Question 8

Research Model and Hypotheses 11

Research Variables 13

Research Organization 14

CHAPTER II. Literature Review 15

Oil Price Shocks and Stock Market 15

Other Macroeconomics and the Stock Market 27

CHAPTER III. Research Model, Data, and Methodology 34

Model 1: The Impact of Oil Revenue on the Stock Market 35

Model 2: The Impact of Gold Price on the Stock Market 36

Model 3: The Impact of Oil Revenue, Gold Price Fluctuation, and Other

Macroeconomic Variables in Stock Market 37

MACROECONOMICS AND THE STOCK MARKET viii

Model 4: The Impact of the Macroeconomic Variables on the Stock Market

Volatility in Iran 38

Methodology 39

Data 40

CHAPTER IV. Empirical Results 42

Variance Inflation Factors 42

Unit Root Test of the Data 43

OLS Results of Model 1: The Impact of Oil Revenue on the Stock Market in Iran 46

OLS Results on Model 2: The Impact of Gold Price on the Stock Market Activity

in Iran 53

OLS Results on Model 3: The Impact of Oil Revenue, Gold Price Fluctuation, and

Other Macroeconomics on the Stock Market activity in Iran 56

OLS Results on Model 4: The Impact of Oil Revenue, Gold Price Fluctuation, and

Other Macroeconomics on the Stock Market Volatility in Iran 61

VAR Results 64

IRF Results 69

VDC Results 76

CHAPTER V. Conclusions 78

Summary of the Study 78

Purpose of the Study 78

Research Questions 79

Research Methods and Models 80

Model 1: The impact of oil revenue on the stock market. 80

MACROECONOMICS AND THE STOCK MARKET ix

Model 2: The impact of the gold price fluctuation on the stock market. 81

Model 3: The impact of oil revenue, the gold price fluctuation, and other

macroeconomic variables on the stock market price index. 82

Model 4: The impact of oil revenue, the gold price fluctuation and other

macroeconomic variables on stock market volatility. 83

Discussion and Analysis of the Research Findings 83

Hypothesis 1: The oil money inflow fluctuation will significantly impact

the Iran stock market index price. 83

Hypothesis 2: The higher gold price will impact the stock market index

price in Iran. 84

Hypothesis 3: Consideration of oil money inflow and gold price

fluctuation in a single model will change the outcome, and the results may

be significantly different. 85

Hypothesis 4: Housing market activity has a direct impact on the stock

market index price in Iran. 85

Hypothesis 5: The foreign exchange rate fluctuation as a parallel

investment market has a significant impact on the stock market index price

in Iran. 86

Hypothesis 6: GDP and the stock market index price are directly related,

and GDP can significantly impact the stock market behavior. 86

Hypothesis 7: CPI has a significant impact on the stock market price in

Iran. 86
MACROECONOMICS AND THE STOCK MARKET x

Hypothesis 8: Interest rate offered by banks has a significant impact on the

stock market activities in Iran. 87

Hypothesis 9: The selected macroeconomic variables have impact on the

stock market volatility in Iran. 87

Future Research Recommendations 90

References 92

APPENDIX 97
MACROECONOMICS AND THE STOCK MARKET xi

List of Tables

Table 1. Descriptive Statistic Data for Global Gold Versus Iran Gold Coin 7

Table 2. Variance Inflation Factor Test Result 43

Table 3. Unit Root Tests of Variables 45

Table 4. Effects of Oil Revenue and Other Macroeconomics on the Stock

Market: Stock(t) = Υ0 + ∑nj=1 Υj macroecon + ξ 47

Table 5. Heteroscedasticity Test: Breusch-Pagan-Godfrey 49

Table 6. Effects of Oil Revenue and Other Macroeconomics on the Stock

Market (adjusted) Stock(t) = Υ0 + ∑nj=1 Υj macroecon + ξ 50

Table 7. Breusch-Godfrey LM Test Results 52

Table 8. Heteroscedasticity Test Results 52

Table 9. The Impact of Gold Price and Other Macroeconomics on the Stock
Market Activity in Iran Stock(t) = φ0 + ∑nj=1 φj macroecon + ωt 53

Table 10. The Impact of Gold Price and Other Macroeconomics on the Stock
Market Activity in Iran Stock(t) = φ0 + ∑nj=1 φj macroecon + ωt.
(Adjusted) 55

Table 11. Heteroscedasticity Test Results on Model 2 (Adjusted) 56

Table 12. The Impact of Oil Revenue, Gold Price Fluctuation, and Other
Macroeconomics on the Stock Market. Stock(t) = Ӯ0 +
n
∑j=1 Ӯj macroecon + Ӫt 57

Table 13. The Impact of Oil Revenue, Gold Price Fluctuation, and Other
Macroeconomics on the Stock Market. Stock(t) = Ӯ0 +
n
∑j=1 Ӯj macroecon + Ӫt 59

Table 14. Heteroscedasticity Test Results on Model 3 (Adjusted) 60

Table 15. The Impact of Oil Revenue, Gold Price Fluctuation, and Other
Macroeconomics on the Volatility of the Stock Market in Iran:
n
Volatility(t) = Ӯ0 + ∑j=1 Ӯj macroecon + Ӫt 62
MACROECONOMICS AND THE STOCK MARKET xii

Table 16. The Impact of Oil Revenue, Gold Price Fluctuation, and Other
Macroeconomics on the Volatility of the Stock Market in Iran:
n
Volatility(t) = Ӯ0 + ∑j=1 Ӯj macroecon + Ӫt (Adjusted) 63

Table 17. Lag Length Selection Criteria 65

Table 18. VAR Result 66

Table 19. VAR Normality Result 69

Table 20. VDC Results 77

MACROECONOMICS AND THE STOCK MARKET xiii

List of Figures

Figure 1. Global Gold Price Versus Iran Gold Coin 8

Figure 2. The Histogram and Jarque-Bera Indicator 48

Figure 3. Examples of Homoscedasticity Versus Heteroscedasticity Distribution

of Data 48

Figure 4. Histogram and Jarque-Bera Test Results (Adjusted) 51

Figure 5. Histogram and Jarque-Bera Test Results 54

Figure 6. Histogram and Jarque-Bera Test Results (Adjusted) 55

Figure 7. Histogram and Normality Test of the Model 3 58

Figure 8. Histogram and Normality Test of the Model 3 (Adjusted) 60

Figure 9. The Histogram and Jarque-Bera Result of the Model 4 63

Figure 10. The Histogram and Jarque-Bera Result of Model 4 64

Figure 11. The Impulse Response of Stock Market to its Own Lags 70

Figure 12. The Impulse Response of the Stock Market to the Oil Revenue 71

Figure 13. The Impulse Response of the Stock Market to the Shock to Housing
Market Activity 72

Figure 14. The Impulse Response of the Stock Market to the Shock to the GDP 73

Figure 15. The Impulse Response of the Stock Market to the Gold Price
Fluctuation 74

Figure 16. The Impulse Response of the Stock Market to the Long-Term Interest
Rate 74

Figure 17. The Impulse Response of the Stock Market to the Foreign Currency
Exchange Rate 75

Figure 18. The Impulse Response of the Stock Market to the CPI 75
MACROECONOMICS AND THE STOCK MARKET 1

CHAPTER I

Introduction

The most important source of energy in the modern world is oil. For this reason, the oil

market continues to be the central focus of attention for economic researchers, and the

relationship between fluctuations in the price of oil and the stock market has been intensively

studied in numerous different countries. The correlation between oil prices and market behavior

has been well documented (Arouri, Lahiani, & Bellala, 2010; Asteriou & Bashmakova, 2013;

Basher, 2014; Bjornland, 2008; Chen, 2010; Dhaoui & Khraief, 2014; Farzanegan, 2008; Le and

Chang, 2011; Kilian & Park, 2007; Mehrara & Sarem, 2009; Sahu, Bandopadhyay, & Mondal,

2013; Oskuee, 2012). These authors, as well as many others, have studied the impact of oil

prices on the stock market at various times in both oil-exporting and oil-importing countries.

Due to the fact that higher oil exports generate larger cash inflow and may strengthen the

economies of the exporting country, the common assumption has been that higher oil prices lead

to higher income and cash inflow to oil-exporting countries, and therefore has a positive impact

on the economy and financial markets of those countries (Arouri et al., 2010; Bjornland, 2008;

Jiménez-Rodríguez & Sanchez, 2004). This is assumed to be true even though higher oil prices

may simultaneously have a negative impact on the economies of oil-importing countries

(Hamilton, 2005; Chen, 2010; Davis & Haltiwanger, 2001). It is further commonly assumed

that, based on the structure of each economy and its dependency on oil, only the time lag and the

intensity of the reaction may vary among different countries. Most of the published research

(Davis & Haltiwanger, 2001; Hamilton, 1983; Kilian, 2009; Park & Ratti, 2007; Sodorsky, 1999)

has focused on developed countries. In most of the oil-exporting countries, stock markets are
MACROECONOMICS AND THE STOCK MARKET 2

still a relatively new phenomenon, and are still developing. Most oil-exporting countries simply

do not yet have a well-established and fully functional stock market.

There are many reasons have been proposed for why stock markets are not yet well-

developed in most of the oil-exporting countries. These reasons include political/government

structure, entrenched hierarchies of the power, lack of social trust, lack of rules to protect

investors, and lack of knowledge among the general public about how the stock market actually

works. Amuzegar (1997) described prerevolutionary Iran as a country that had “completed

nearly two decades of rapid economic growth and impressive social development” (p. 4). He

explained how Iran was transformed from an “agrarian and stagnant” (Amuzegar, 1997, p. 4)

economy into a modern progressive economy. Amuzegar (1997) referred to the data from World

Bank showing that

compared with other Third World countries, Iran’s annual real growth rate of nearly 9.6

percent during 1960-1977 was roughly double the average of the countries in the middle-

income category, and higher than the average for any other group of countries in the

world. (p. )

A comparison of the U.S. dollar exchange rate against the Iranian rial between 1979 and

the present rate (from about 75 rial per dollar to 38,500 rial per dollar) as well as a substantial

dynamic inflation shows the drastic depreciation of rial.1 This depreciation reduces the power of

the domestic currency each year, and, as a result, most ordinary Iranian citizens convert their

surplus money into a more stable asset, primarily gold, real estate, or foreign currency (U.S.

dollars) to hedge their money against inflation. In most well-developed stock markets, gold

mines and gold producers ordinarily have their stock available to the public for investment, but

in Iran, the gold market is a totally separate market. Because people traditionally consider gold
1
Data are from the Central Bank of Iran (n.d.).
MACROECONOMICS AND THE STOCK MARKET 3

as one of the safest investments and hedge tools, this gold market plays a substitute role parallel

to the stock market. In other words, gold is not a part of the Iranian stock market, but rather

serves as an alternative to the stock market for investing and hedging purposes. This relationship

is totally different from that which exists in the stock markets of the developed Western

countries. Real estate is another popular investment market and has the same relationship to the

Iranian stock market as does gold in comparison with the Western countries stock market. While

investors in the United States can invest in real estate using real estate investment trusts (REITs)

and mutual funds, there is no analogous opportunity in the Iranian stock market, so real estate

also can be considered as an alternative investment tool to the stock market. Therefore, for the

purposes of this paper, gold, real estate, and the foreign exchange currency market are

considered to be parallel and substitute markets to the stock market, and the relationship between

variables will be investigated from this point of view.

Background of the Problem

As a developing country, Iran has a relatively active stock market compared to the other

Middle Eastern countries. According to the Tehran Stock Exchange (TSE) website

(new.tse.ir/en), this market was established on 1967 with six companies and a total capital of 6.2

billion Iran rials. Its current value is estimated to be approximately 2,789 billion rials. This

history as well as the history of the economy and financial activity in Iran as a whole can be

divided to three distinct periods: (a) the pre-revolution era from 1967 to the revolution on 1978,

(b) from the revolution to end of Iran-Iraq War in 1988, and (c) from the end of the Iran-Iraq

War to the present. The economy of Iran started moving towards modernization in about 1960,

and soon after this point began expanding dramatically. Unfortunately, this momentum was

completely halted by the onset of the revolution in 1979. Almost as soon as the revolution was
MACROECONOMICS AND THE STOCK MARKET 4

successful, the country was almost immediately drawn into the war with Iraq, which lasted

another 8 years, during which time the economy experienced the unique situation of what is

referred to as the “war economy,” during which time there was essentially no real economic

growth. Although the Iranian economy had experienced some growth and modernization during

the 1960s and 1970s, the greatest level of stock market activity did not really begin until after the

end of the Iraq-Iran war in the late 1980s (TSE, n.d.). However, even though the market volume

and activity have risen drastically, this cannot be considered as an indicator of investment and

financial activity for the entire population of Iran. Among ordinary investors in Iran, the stock

market is considered to be a very new and unfamiliar investment method, which is neither well-

understood nor trusted by the general public. Guiso, Sapienza, and Zingales (2008) studied the

relationship between a general lack of trust among the public and the stock market and

concluded that this lack of trust among individuals (and within Iranian society at large) is one of

the prime factors that prevents potential investors from investing in the stock market. Among

other reasons claimed by Guiso, the widespread fear of being cheated and the general lack of

public confidence in the existing investor protection rules leads most ordinary Iranian citizens to

invest in gold, real estate, and other tangible assets instead of investing in the stock market.

Although it is common knowledge that the economy of Iran is intimately tied to the production

and sale of oil, and it is generally assumed the nation’s stock market must be greatly influenced

by the inflow of income from the sale of oil, very few studies have actually been published about

the reaction of the Iranian stock market to fluctuations in the price of oil. There are also studies

available about the relationship between the oil prices and other macroeconomic variables in Iran

(Mehrara, 2006; Monjazeb & Hoseinzadeh, 2013; Rad, 2011). However, there does not seem to

have been any studies which investigated the relationship between the three most popular
MACROECONOMICS AND THE STOCK MARKET 5

investment markets (gold, real estate, foreign currencies) and the stock market nor do there seem

to be any available studies which examine the reaction of these other markets to the fluctuations

of oil revenue from the perspective of Iranian investors.

Purpose of the Study

Because Iran is an oil-exporting country whose economy is almost completely dependent

upon the inflow of oil export money, much attention has been given by researchers to the impact

of oil price fluctuations on various macroeconomic sectors of the Iranian economy during recent

years. A study by Oskooe (2012) concluded that

in view of underlying data-generating process of the series, the variance of oil price

fluctuations does not cause the variance of Iran stock returns. This means that there is no

volatility spillover effect between Iran[’s] stock market and [the] international oil market.

(p. 450)

It seems that Iran is taking steps to create a better relationship with the rest of the world

generally, and with the Western nations in particular.2 If this could be accomplished, it would

greatly advance the economic development of Iran generally, and would (presumably) have a

great impact upon the Iranian stock market. Meanwhile, the huge young adult populations of

Iran, many of whom are college graduates, have a much better understanding of the stock market

than did the previous generation, would be more likely to invest in the stock market, and, by their

participation in it, press for improvements in the way it operates.

Furthermore, the recent dramatic expansion of the digital information infrastructure

throughout Iran has greatly increased the flow of knowledge and dissemination of information

2
The agreement on the nuclear activity between Iran new government and the world (under the leadership of the United States) is considered as a
diplomatic success in the Iranian foreign policy and can be considered as a signal of taking steps toward a better relationship with the Western
world. Iranian officials have also taken the position to change the relationship with the western world recently. For example, see the following:
https://www.theguardian.com/world/2013/jun/17/iran-hassan-rouhani-promises-moderation
http://www.al-monitor.com/pulse/originals/2015/10/iran-west-relations.html
http://www.ft.com/cms/s/0/01261236-da16-11e4-ab32-00144feab7de.html
MACROECONOMICS AND THE STOCK MARKET 6

about the stock market throughout the general population of the country. It may be expected that

this great increase in the availability of information about the stock market can be expected to

make the stock market at least as a considerable option in the future for Iranian investors.

The main purpose of this study is to arrive at a clear picture of the behavior of the Iranian

stock market, its relationship to the oil market and other available investment opportunities, and

its role as an investment tool for Iranian investors in the future.

Importance of the Study

The stock market is considered to be one of the most important components of any

modern economy, which provides the access to the capital of the public and makes it available

for use by business entities. If we assume that the economic relationship between Iran and the

United States would be stronger in the future and the Iranian government would be more

successful in improving its relationship with the western world, in general terms, as an integral

part of a developing economy the stock market of Iran can be expected to be a more important

investment method in the future, and will be expanded as it absorbs an increasing amount of the

investment capital from domestic investors. It is vital for the policy makers who are responsible

for planning the development of the economy to educate themselves about this subject, and to

have a broad, general understanding of the behavior of this market. Furthermore, because gold is

one of the most popular tools the Iranian government uses to implement its monetary policy, it is

very important for monetary authorities.to have a clear understanding of the possible

relationships between gold and the stock market. This study may also become a beginning step

for further research into the behavior of the stock market. For example, the numerous social,

cultural, and psychological factors operating within Iran that lead individuals (or the whole
MACROECONOMICS AND THE STOCK MARKET 7

society) to decide to invest in stocks rather than in gold or real estate has never been clarified and

considerably more research is needed into these subject.

Scope of the Study

The revolution of 1979 was soon followed by invasion from Iraq in 1980, which was in

turn followed by a series of severe economic sanctions against Iran by United States and other

nations. This series of events, have made the economic history of Iran quite unique and

unusually sensitive to global financial events and internal political developments. As a result,

Iranian investment markets, including gold, real estate, and the stock market frequently seems to

overreact to political and economic news. For example, in response to internal political events, it

is very common that the degree of fluctuation in the price of gold within Iran is greater or lower

than the degree of fluctuation of the global gold price. By the other words, the investment

market in Iran is extremely influenced by both the internal and the external (global) political

events. Table 1 indicates the descriptive statistic of annual price of global gold as well as the

Iran annual gold coin followed by their graph for the period of 1995 to 2012.

Table 1

Descriptive Statistic Data for Global Gold Versus Iran Gold Coin

Global Gold 1000 USD Iran Gold Coin 1M Iran Rial

Mean 0.631173 1.859389
Median 0.39852 0.834
Maximum 1.66898 10.2
Minimum 0.27104 0.385
Std. Dev. 0.450729 2.47495
Skewness 1.264685 2.454829
Kurtosis 3.286996 8.437815
MACROECONOMICS AND THE STOCK MARKET 8

.6

.5

.4

.3

.2

.1

.0

-.1

-.2
1996 1998 2000 2002 2004 2006 2008 2010 2012

Global Gold Iran Gold Coin

Figure 1. Global gold price versus Iran gold coin.

As it appears, the standard deviation of the Iran gold coin (SD = 2.475) is a significantly

higher than the standard deviation of the global gold price (SD = 0.451). This indicates that the

degree of fluctuation of the Iranian coin has been higher than the global gold price. This may

sometimes make the relationship between the macroeconomic variables more complicated, for it

is difficult (if not impossible) to determine whether the fluctuation is due to a global price change

or to some political issue that is purely internal to Iran. Also relevant to the scope of this study is

that, due to the general absence foreign investors within Iran and a widespread lack of

knowledge about investment in the stock market on the part of domestic investors, it can be

difficult to generalize the findings of the study (or any study) to all investors and all situations

within the country.

Research Question

Due to the nature of the “oil based economy” as well as cultural barriers which were

discussed earlier in this chapter (Guiso et al., 2008), and based on the quantity of the body of
MACROECONOMICS AND THE STOCK MARKET 9

research available in the literature, it appears that in most of the oil-exporting countries, stock

markets are a relatively new and developing institution, and their relationship with other

financial institutions and economic variables in most of these countries still requires considerably

more research. Although there have been a few studies conducted and published about oil-

exporting economies and stock markets, many of the most important questions, issues and

variables still remain largely unexplored.

For the purpose of this study, several major assumptions and considerations about the

Iranian economy will frame the investigation. First of all, it must be understood that the

economic structure and market system of Iran, as well as pattern of relationships among key

macroeconomic variables, is unique to this country and therefore must be studied with the

cultural, historical, and political environment of Iran in mind. Second, the must be kept in mind

that the high level of corruption within the government and the entrenched hierarchy of the

political power in Iran have created an inordinate lack of trust and high level of fear of being

cheated within potential investors. Third, it must be kept in mind that high inflation rapidly

reduces the value of the domestic currency (Iranian rial), and therefore that any investment will

not be perceived to be satisfactory, regardless of its rate of return, unless it has historical been a

hedge against inflation. Fourth, as the result of the first three considerations, the investment

process in Iran has not yet been modernized (specifically in the form of portfolios). Also, due to

the general depreciation of Iran rial, the real rate of return of the stock market is not beneficial

and in most cases, ordinary Iranian citizens prefer to invest in tangible assets such as gold, real

estate, or foreign currency, which satisfy their desire for a hedge against inflation as well as

providing an acceptable rate of return. Fifth, those tangible assets (gold, real estate, and foreign

currency) have become an alternative investment option to the stock market. In other words, as
MACROECONOMICS AND THE STOCK MARKET 10

mentioned previously, the purchase of gold stock and real estate stock (such as REITs), which is

possible in the American or European stock markets, are not investment options in the Iranian

stock market. As a result, each of these markets work together as parallel markets, and behavior

of each of these markets or combination of these markets has an impact upon all the others.

Because some previous studies (Oskooe, 2012) reported that oil price fluctuation has no

significant impact on the Iranian stock market, this study examined the following questions.

1. Does the oil money inflow3 have any impact on the Iranian stock market price index?

2. Is there any significant relationship between the gold market and the stock market in

Iran?

3. Is there any significant relationship between the real estate market and the stock

market index in Iran?

4. Is there any significant relationship between the long-term deposit interest rate paid

by banks and the Iranian stock market?

5. Is there any significant relationship between the foreign currency (U.S. dollar)

exchange rate and the Iranian stock market price index?

6. Would the result of the study be different if it took into consideration all of the above

variables in one single model?

The main goal of this study is to focus on the impact of oil and the gold price fluctuation

on the stock market in Iran. This is because the former is the main source of money inflow in

Iran economy, and the latter is the primary investment and hedge market traditionally used by

Iranian investors. There are several reasons that the gold market is more important than other

investment markets. For example, gold is an extremely liquid asset. The current value of the

3
The term oil money inflow is used in this study to represent Iranian oil revenue instead of oil price fluctuation.
This is because Iranian oil income might be influenced by political issues such as global sanctions, regardless of the
global price of oil.
MACROECONOMICS AND THE STOCK MARKET 11

investment can be determined quickly and easily by an investor (or potential investor), which

means that any investor with any budget can invest confidently in the gold market. Furthermore,

as mentioned previous, an investment in gold is a hedge against inflation.

This study assumes that increased oil money income coming into Iran will have a positive

impact on the general financial activity within the country. The main question being investigated

is whether the increased oil income, when considered along with other economic variable

indicators, can influence the stock market price index in Iran.

Research Model and Hypotheses

The impact of oil price fluctuations on Iranian macroeconomic variables and the Iranian

stock market has been well-documented, and several studies have been published (Farzanegan,

2011; Oskooe, 2012). Even so, there has been no investigation about the combined impact of oil

money income and gold price upon the stock market. Because gold is considered to be one of

the most important investing and hedging tools in Iran, this study is aims to discover whether any

relationship exists among all of the above variables. To achieve this goal, this study develops

three different models to evaluate the possible relationship between oil, gold, and other

macroeconomics on the Iran stock market separately. When developing macroeconomic models,

exogenous variables are variables that affect the model without being affected by it. Such

exogenous variables are those which are assumed to be completely controlled by factors beyond

the scope of the model. Conversely, endogenous variables are those variables whose outcome is

generated based on the relationship between other variables within the model. For the purposes

of this study, the Iranian Stock Index is considered to be an endogenous variable, while all of the

other macroeconomic variables are considered to be exogenous variables. All the models

employed in this study assume the stock market price index to be dependent variable to other
MACROECONOMICS AND THE STOCK MARKET 12

independent macroeconomic variables such as oil income, gold price, housing activities,

consumer price index (CPI), and gross domestic products (GDPs). Based on the vector auto

regression (VAR) model designed for this study, the impact of oil money inflow will first be

evaluated. After this, the impact of CPI, housing activity, foreign exchange rate, and GDP on the

stock market will be evaluated. Next, ceteris paribus, gold price will be replaced with oil money

income in the model. Finally, gold price and oil income will be added together into the equation,

and all results will be compared as the conclusion of the study. Previously research on the topic

(Oskooe, 2012) does not claim that despite the severity of the dependence of the Iranian

economy upon oil money, fluctuation in the volume of oil exports nor do fluctuations in the price

of that exported oil have any significant impact on the Iranian stock market. This study intends

to retest this claim as a part of an investigation into the relationship between these two variables.

As mentioned previously, the price of gold and the level of gold market activity may impact the

stock market, so therefore, as another part of the investigation this study will test this relationship

as well.

Based on the above assumptions, the hypotheses of this study are as follows:

1. The fluctuation in oil money inflow does significantly impact the Iranian stock

market index price.

2. An increase or decrease of the price of the gold does impact the stock market index

price in Iran.

3. Consideration of oil money inflow and gold price fluctuation together within a single

model will change the result of the test and the result may be significantly different.

4. Housing market activity has a direct impact on the stock market index price in Iran.
MACROECONOMICS AND THE STOCK MARKET 13

5. The foreign exchange rate fluctuation, as an investing option for investors in Iran,

will have a significant impact on the stock market index price in Iran.

6. GDP and the stock market index price are directly related, and GDP can significantly

impact the stock market behavior.

7. CPI has significant impact on the stock market price in Iran.

8. The interest rate offered by banks has a significant impact on the stock market

activities in Iran.

As mentioned previously, this study is designed to test the first and the second

hypotheses separately in combination with other variables. Each outcome will be discussed

separately. At the conclusion, this paper will also run the same general model for stock market

volatility to examine any relationship between the macroeconomic variables and the stock

market volatility.

Research Variables

This section delineates and defines the variables used in this study as follows:

1. Stock: Represents the logarithm of the second difference of the TSE price index data

for time t.

2. Gold: Represents the logarithm of the second difference of the price of gold coin

issued by Central Bank of Iran for time t.

3. Oil: Represents the logarithm of the second difference of oil revenue during the

period of the study.

4. House: Represents the logarithm of the second difference of housing market activity

in Iran during the period of the study.

MACROECONOMICS AND THE STOCK MARKET 14

5. GDP: Is the logarithm of the second difference of gross domestic products of the

country for time t during the study.

6. CPI: Stands for logarithm of the second difference of consumer price index as an

indicator of price level during the study.

7. Exchange: Represents the logarithm of the second difference of real unofficial

exchange rate of the Iranian rial for the U.S. dollar.

8. Interest: Represents the interest rate offered by banks for five year deposits.

9. Volatility: Represents the stock market annual standard deviation as a proxy to

measure the market volatility. The data is then transformed into quarterly series to be

fit to the model.

Research Organization

The remainder of this research will be organized as follows: Chapter 2 will review the

literature relevant to the topic of the study. This chapter will review the literature about the

impact of each variable on the stock market in various different countries. Research design and

methodology will be presented in Chapter 3, followed by data sources and analysis. The models

are also discussed in detail within this chapter. The result and outcome of the test is provided in

Chapter 4. Finally, in the conclusion, analysis of the outcome, recommendations for future

researchers and policy makers will be presented in Chapter 5.

MACROECONOMICS AND THE STOCK MARKET 15

CHAPTER II

Literature Review

This chapter reviews the literature regarding the impact of the critical variables on the

stock market in several different countries. This part of the study is organized as three steps.

First, the literature about the impact of oil price fluctuations on the stock market in different

countries is reviewed. After this, literature about the impact of the gold price on the stock

market will be examined. Finally, as the third step, the literature about the impact of other

variables upon the stock market, including housing activity, GDP, CPI, and foreign exchange

rate, will be reviewed.

Oil Price Shocks and Stock Market

Petroleum is one of the most important sources of energy, and therefore the entire global

economy is closely tied to its production and sale. Fluctuations in the price of oil have therefore

been extensively studied for decades. The influence of the changing price of oil has been very

well documented. The consequences of oil price fluctuations and their impact on financial

markets and macroeconomic variables has been monitored and analyzed in numerous countries

from a wide diversity of perspectives. Due to the severity of the impact of oil prices on the

global economy, the shocks, causes, and effects of price fluctuation have been the center of

attention for many researchers and economists. Hamilton (2005) claimed that nine out of 10 of

the U.S. recessions after World War II were due to a positive shock to the oil price. Hamilton

developed an ordinary least square (OLS) model based on the lagged change in GDP growth

rates and lagged logarithmic change in oil prices, and inferred that the relationship cannot be a

statistical coincidence. Kilian (2009) said that all of the shocks to the oil market cannot be

placed in the same category, and as a result, the impact of each shock on the real oil price may be
MACROECONOMICS AND THE STOCK MARKET 16

different. Based on the data from the United States, Killan divided the shocks to the oil market

into three categories. These three categories were: (a) crude oil supply shocks (possible political,

economic, or technical issues in oil-exporting countries), (b) shocks to the global demand for all

industrial commodities, and (c) demand shocks that are specific to the crude oil market (mostly

precautionary demand, such as anticipation of cold weather causing concerns about possible

future reductions in oil production).

Using a VAR model, Killan (2009) has designed a model based on the impact of different

kinds of shocks on the price of the oil. In this model, A0zt = α +Σ Aizt-I + ε, where έ denotes the

vector of serially uncorrelated structural innovations and zt = (∆prodt, reat, rpot). In this model,

prod is defined as percentage change in oil production, rea as real economic activity, and rpo as

the real price of oil. Killan also stated that ȩ=A0έt where έ would represent the oil supply shock,

aggregate demand shock, and oil-specific demand shock separately. Killian interpreted the result

of the outcome of his model by stating that an increase in precautionary demand for oil has an

immediate, persistent, and large increase in the price of the oil, while an increase in aggregate

demand for all industrial commodities has a delayed and sustained increase in the price of oil.

Killian further stated that oil production disruptions have a small and transitory increase in the

price of oil. He then concluded that traditional thinking that is found in most conventional

economics literature which attempts to analyze the effects of higher oil prices on the U.S.

economy is not well defined. According to Killan, this is due to the existence of a reverse

causality from macroeconomic aggregates to the oil price, and also due to the fact that the

structural demand and supply shocks that drive oil prices have both direct and indirect effect on

the U.S. economy as well.

MACROECONOMICS AND THE STOCK MARKET 17

In another study similar to Kilian (2009), Basher (2014) also arrived at the same

conclusion for three groups of oil-exporting, oil-importing, and emerging countries. However,

Davis and Haltiwanger (2001) investigated the impact of oil price fluctuations on job creation

and destruction in the United States during 1972 to 1988. Like Killian, they also used VAR

analysis to develop a model which used the oil price index, its absolute change, the

manufacturing job creation and destruction rate, the sectorial job creation and destruction rate,

and the interest rate quality spread as variables. The outcome of the OLS model result led to the

conclusion that oil shocks account for 18 to 25% (about twice as much as monetary shocks) of

the variance in employment growth, as well as to much of the job reallocation response to oil

shocks which occurs within energy-intensity categories. Davis and Haltiwanger (2001) also

concluded that there are short-term employment declines in response to an unanticipated increase

in the price of oil. Consistent with previous studies, they also found that employment growth is

impacted by an increase in the price of oil considerably more sharply than it is impacted by the

decrease in the price of oil. A positive shock to the oil price result was a 2% decline in

employment growth, which was about ten times as strong as the reaction to negative shock of the

same magnitude to the oil price.

Chen (2010) investigated whether a higher oil price pushes the stock market into bear

territory. In his study, instead of focusing on the impact of oil prices on stock market returns, he

focused on the stock market behavior and changes from bull to bear market and vice versa as a

reaction to the oil price. Based on the monthly data from the period of 1957 to 2009 of the S&P

500 and average global oil price, Chen found that an increase in oil price raises the probability of

switching from a bull market to a bear market, and that a higher oil price causes the market to

stay in a bear condition for a longer period.

MACROECONOMICS AND THE STOCK MARKET 18

Park and Ratti (2007) examined oil price shock and the stock market in the United States

and 13 European countries (Germany, Spain, the United Kingdom, Italy, Norway, Austria,

Belgium, Sweden, Finland, Netherlands, France, Greece, and Denmark), for the period of 1986

to 2005. They used short-term interest rate (r), real oil price (op), industrial production (ip), and

real stock returns (rsr) as variables to develop the model as: VAR (r, op, ip, rsr) for each

country. By using the impulse response function, they reported that a one standard deviation

shock to the oil price has a statistically significant negative impact on the stock market returns of

the United States and 10 of the 13 European countries (the exceptions were Norway, the United

Kingdom, and Finland) at the 5% confidence level, either in same month or the month

immediately following the shock. Park and Ratti also estimated the impact of the volatility of

the oil price on the other variables in the model and inferred that for many European countries,

but not for the United States, the volatility of the oil price and the uncertainty of the future

market would depress the real stock return during the same month or the month immediately

following the shock to the price of oil. The median of variance of the outcome showed that oil

price shocks account for 6% of the volatility in the stock market (the range went from 3% for the

United Kingdom to more than 10% for Sweden). These results also showed that a shock to the

oil price would significantly increase the short-term interest rate in the United States and eight

other European countries (Park and Ratti described the phenomena as tightening the monetary

policy to prevent the inflation as a result of the shock).

Cologni and Manera (2005) investigated whether the oil price shocks have been

transmitted to the monetary policy actions globally by investigating the oil price shock impact

directly on the interest rate and inflation for G7 countries (Canada, France, Germany, Italy,

Japan, the United Kingdom, and the United States). They used short-term interest rate, monetary
MACROECONOMICS AND THE STOCK MARKET 19

aggregate, CPI, real GDP, world price of oil, and exchange rate (as the ratio of SDR to the U.S.

dollar) for each country, then developed a VAR model to estimate the relationship between oil

price and inflation as well as the interest rate for those countries. Cologni and Manera found that

an unexpected shock to the oil price is followed by an increase in the inflation rate and also

reduces the output growth for most of the countries in the study. They also concluded that

authorities within most countries of the study reacted to the oil price shock by increasing the

interest rate. This was done as an attempt to fight the possible inflation that might result from

the shock. Asteriou and Bashmakova (2009) investigated the relationship between the oil price

risk and the stock market for the capital markets of central European and eastern European

countries, including the Czech Republic, Estonia, Hungary, Lithuania, Latvia, Romania, Poland,

Russia, Slovakia, and Slovenia from 1999 to 2007. They used the capital asset pricing model to

estimate the beta of the oil market by using Rit as daily excess return for each country at time t,

Rmt as daily returns on the market index, Rot as return on oil prices, Ret as daily exchange rate

return, and then developed their model as Rit=α+β1, itRmt+β2, itRot+β3, itRet+έe,t. The test results

revealed that the coefficient of the market beta was statistically significant and positive for all

countries within the study. This indicates the direct relationship between market risk and stock

market return. They also found that the oil price risk has a statistically significant negative

coefficient, which suggests that an increase in oil price returns causes decreases in stock market

returns for all countries included in the study. They concluded that the oil price fluctuation had

the most significant impact on the stock markets of central and eastern European countries.

Sahu et al. (2013) also examined the relationship between oil price and the stock market

in India, based on the daily data from 2001 to 2013. They used the vector error correction

model, which calculates and ignores the cointegration between two or more variables in the
MACROECONOMICS AND THE STOCK MARKET 20

model by turning all variables to first difference I(I) instead of the Level I(0). The vector error

correction model is broadly used to estimate the long-term relationship between the variables.

Sahu et al. concluded that the price of crude oil has no significant causal effect on the Indian

stock market.

Because Norway is an oil-exporting country, the Norway stock market is also heavily

influenced by the global oil price shocks. Bjornland (2008) investigated the response of the

Norwegian stock market to positive shocks to the oil price. Like most others who use time series

to investigate the impact of oil price on macroeconomic variables and financial markets,

Bjornland also used a VAR model, and selected oil price, stock price, domestic and foreign

interest rate, unemployment rate, domestic CPI change as inflation, and real exchange rate as

variables. Bjornland then used the monthly data for the period from 1993 to 2005 to developed

his model. In this model, Zt = α + ∑ AiZt-I + Ʋt , where Zt is the vector of endogenous

variables, α is the intercept vector, Ai is the ith matrix of autoregressive coefficients and vt is a

vector of reduced form white noise residuals. Based on the outcome of the model, Bjornland

inferred that higher oil price increase the stock return in Norway. He concluded that following a

ten percent increase in the oil price, stock returns in Norway immediately increase between 2 and

3%. He reported that the maximum impact of an increase in oil price is reached about 14 to 15

months after the shock, with approximately a 4 to 5% increase in return. According to

Bjornland, after about 14 to 15 months, the effect dies out. In general, his work indicates that the

Norwegian economy responds to higher oil prices by increasing aggregate wealth and demand,

which in turn tends to reduce the unemployment rate and increase the inflation rate and interest

rate.
MACROECONOMICS AND THE STOCK MARKET 21

Sadorsky (1999) used the natural logarithm of the U.S. industrial production as a proxy

for output, interest rate, and real oil price to develop a VAR model. He tested not only oil price

fluctuation but also the effect of oil price volatility on the U.S. stock market. He concluded that

both the price and its volatility, (which creates uncertainty) play important role to affect

economic activity. He concluded that a shock to the oil price would depress the stock market

return in the United States, and that a shock to the stock market has positive impact on both the

interest rate and upon industrial production.

In another study, Jiménez-Rodríguez and Sanchez (2004) examined the impact of oil

price shock on the GDP growth of several countries within the Organization for Economic

Cooperation and Development, including Canada, France, Italy, Germany, Japan, Norway, the

United Kingdom, the United States, and Euro Area as whole, using data from 1970 to 2001.

These authors also used real GDP, nominal oil price, short and long term interest rate, real

effective exchange rate, real wage, and CPI as seven variables to develop their VAR model. In

this study, the outcome of Granger causality tests indicated that oil price fluctuation and

macroeconomic variables interact at least in one direction (either positive or negative) for all

countries, and in two directions (positive and negative) in most countries. These findings were

consistent with other literature about cases of asymmetric reaction of various economies to oil

price increases and decreases in which oil price increases have greater impact on the economy

than do oil price decreases. They also found that while Norway as an exporter benefits the raise

in the oil price, the U.K. GDP growth reacts significantly negative to the price rise. The authors

traced this to the sharper real exchange-rate appreciation in the United Kingdom. On the other

hand, as another anomaly, Jiménez-Rodríguez and Sanchez reported that GDP growth in Japan

(an oil-importing country), reacted positively to the oil price rise. The authors were unable to
MACROECONOMICS AND THE STOCK MARKET 22

offer any explanation for this counterintuitive observation, and simply suggested that some

peculiar (and unspecified) set of circumstances operating within the Japanese economy must

have been exerting an influence during the period of the study.

In another study, Basher and Sodorsky (2006) focused on the stock market within

emerging economies. They used an international multi-factor model to estimate both conditional

and unconditional risk factors, which was based on the capital asset pricing model. They used

the daily closing price of 21 emerging economies and the Morgan Stanly Capital International

World Index for the period of December 31, 1992 to October 31, 2005. They used data for

emerging economies, including Argentina, Brazil, Colombia, Chile, India, Indonesia, Israel,

Jordan, Korea, Malaysia, Mexico, Pakistan, Peru, Philippines, Poland, South Africa, Sri Lanka,

Taiwan, Thailand, Turkey, and Venezuela. Basher and Sodorsky concluded that there is a

significant negative relationship between oil market beta and emerging stock market pricing.

They also concluded that there is a positive relationship between beta and return when the

market is up, but a negative relationship when the market is down. This finding is consistent with

other published research from this era.

Maghyereh (2004) examined the oil prices and stock markets of 22 emerging countries

by using the VAR model based on the daily data for the period from January 1998 to April 2004.

Using the variance decomposition (VDC) analysis, which estimates direct impact of each

variable on the fluctuation of the other, they concluded that, in contrast with countries with

developed markets, an oil price shock does not have a significant impact on the stock market of

these emerging countries.

In another study, using data from 1999 to 2011, Wang, Wu, and Yang (2012) analyzed

the relationship between oil prices and stock return in oil-importing and oil-exporting countries
MACROECONOMICS AND THE STOCK MARKET 23

separately. For oil importers, they used data from the United States, Japan, Germany, France,

United Kingdom, Italy, China, Korea, and India as oil importers. For oil exporters, the used data

from Saudi Arabia, Kuwait, Mexico, Norway, Russia, Venezuela, and Canada. Wang et al. also

used VAR model to analyze their data, and concluded that the magnitude, duration, and even

direction of the impact is very different in oil-exporting and oil-importing countries. They also

concluded that, in general, oil price shocks explain approximately 20 to 30% of global stock

return. The authors also claim that the uncertainty in the supply of oil can depress the stock

markets of both oil-importer and oil-exporter countries.

Le and Chang (2011) also studied the reaction of the stock market to the oil price

volatilities in Japan, Malaysia, Singapore, and Korea from 1986 to 2011 using impulse response

and VDC function. They analyzed two different hypotheses as positive and negative relationship

between the oil price and the stock market behavior. In this study, the positive relationship

hypothesis assumed that rising oil prices are associated with a booming economy and stronger

businesses performance, and result in higher demand for oil. From this perspective, there is a

mutual relationship between oil prices and stock market prices. In contrast, the negative

relationship hypothesis assumes that oil prices and stock market prices have a one-way

relationship, in which changes in the price of oil impact the stock market, but that changes in the

stock market activities do not impact the price of oil. Le and Chang concluded that the impact of

oil prices on the stock market is not statistically significant in the countries included within their

study. They also claimed that in an inefficient stock market, the reaction of economy and stock

market activity to an oil price shock is slow. They make this claim because, based on their

model and their data, it took a fair amount of time for stock markets of Singapore, Korea, and
MACROECONOMICS AND THE STOCK MARKET 24

Malaysia to react to the oil price shock (inefficient markets), while the Japanese stock market

reacted immediately (efficient market).

There are very few studies available on the financial markets of the oil-exporting

countries of the Middle Eastern. As one on those few studies, Mehrara and Sarem (2009) tested

the effect of oil price shock on the industrial production of Iran, Saudi Arabia, and Indonesia

during the period from1970 to 2005. They used a Gregory-Hansen cointegration analysis and a

Granger causality test to examine real oil price and the industrial output as macroeconomic

variables. They used real industrial value added as a proxy for measuring output in the traded

goods sector and found that the result of the Granger test is qualitatively the same for Iran and

Saudi Arabia. This was because the coefficient of the error correction term is significant in the

energy equation. This indicates that long term causality runs from oil price to the output for both

countries, and this implies that the two countries are highly vulnerable to external shocks,

particularly to oil price decreases. In contrast, for the case of Indonesia, they found no

cointegration relationship between oil price and output. This indicates that oil shocks have only

small effects on the output in this country; which the authors interpreted it as meaning the

efficiency of natural resource trends to promote economic diversification and growth within

Indonesia.

Arouri et al. (2010) estimated the relationship between oil price shocks and stock markets

in the Gulf Cooperation Council countries, all of which are Middle Eastern, oil-exporting

countries, which include Bahrain, Oman, Kuwait, Qatar, Saudi Arabia, and United Arab

Emirates. The authors designed both linear and non-linear models to estimate the short-term and

long-term impact of oil price fluctuations on the stock market for each of the countries within the

study. They found that oil price shocks significantly impacted the stock market in Qatar, Oman,
MACROECONOMICS AND THE STOCK MARKET 25

Saudi Arabia, and the United Arab Emirates, but apparently had no significant impact on stock

markets of Kuwait and Bahrain.

Oskooe (2012) tested the causality in mean and variance to investigate the relationship

between oil price shock and the stock market in Iran. He used weekly data from 1999 to 2010 to

test the causality of first and second moment dynamics of the oil price and the Iranian stock

market data series. By comparing their data set from the oil price and the Iranian stock market,

he noticed that fluctuation in oil prices have been greater than Iranian stock prices. He also

reported the price of oil has a greater standard deviation compared to the Iranian stock market,

which indicates the high volatility and riskiness of the oil market. He interpreted the positive

value of skewness and kurtosis of the data as a higher frequency of possibility of large positive

return than a large negative return in the Iranian stock market. He found that there is a positive

weak correlation between the Iranian stock returns and oil price fluctuations. He then used an

autoregressive integrated moving average and a generalized autoregressive conditional

heteroscedasticity model to analyze his data. He finally concluded that there is no volatility

spillover between oil price and the Iranian stock market. This indicates that information related

to fluctuation in oil price does not affect the volatility of Iran stock market. Oskooe interpreted

this as a positive sign to investors in the Iranian stock market. He did so because the lack of

volatility spillover between oil price shocks and the Iranian stock market is generally taken as a

sign of invulnerability of this market against possible external shocks to the oil price.

Farzanegan (2011) investigated Iranian government behavior regarding the spending oil

revenue. Because the Iranian economy is almost totally dependent on oil revenue, the spending

behavior of the government is very important for the entire national economy, and can exert a

powerful and direct impact on businesses and the stock market. He has used VDC and impulse
MACROECONOMICS AND THE STOCK MARKET 26

response functions (IRFs) with data from 1959 to 2007. Farzanegan categorized government

spending in Iran into two groups: (a) current expenditure (goods and services, social security and

pensions, interest payments, subsidies, military, health, education, cultural, and social activities),

and (b) capital expenditure (investments by the government in infrastructures and public goods).

He argues that a rise in the oil price would increase the size of the government by increasing

current expenditures; but the government is not able to reduce its size and activity in the case of

decreasing of the oil price. Therefore, he says the result of this pattern is higher budget deficits.

Based on the outcome of his models, only government expenditure on the military sector show

any significant response to the oil price increases, while none of the other variables react

significantly to the oil price shock.

In a separate study, Farzanegan and Markwardt (2007) also investigated the impact of oil

price shock on the Iranian economy. They used real industrial GDP per capita, real public

consumption, imports, exchange rate, inflation, and oil price change as macroeconomic variables

to develop their VAR model. This model stated that yt = A1yt + …. + Apyt-p + Bzt + έt and

defined yt as endogenous variables, zt as vector of exogenous variables, Ai and B as coefficient

matrices and p as lag length. The authors used the IRF tool to examine the effects of oil price

shocks on the Iranian macroeconomy. They also used VDC technique to measure the amount of

impact of each variable on the others. They found that positive oil price shock accounted for 1 to

2% of inflation, while negative oil price shock explained about 5 to 8% of inflation. These

findings demonstrate the severe impact of financing budget deficit during a negative market on

inflation. As the outcome of the study, Farzanegan and Markwardt concluded that positive oil

price shock increases the effective exchange rate and appreciates the Iranian rial against the U.S.

dollar, and it causes the “Dutch disease” phenomena in the economy. This is because imports
MACROECONOMICS AND THE STOCK MARKET 27

become cheaper, exports become more profitable, and the whole process raises inflation

significantly. Furthermore, based on their findings, they concluded that the Iranian rial

depreciates in response to a negative oil price shock, and that this can be another source of

vulnerability of the Iranian economy.

Farzanegan and Markwardt (2007) highlight the importance of having a stabilizing fund

to control the effects of oil price fluctuation on the economy. They did not consider stock market

behavior in reaction to the oil price in their study (the stock market was not one of the variables

in their study). Even so, the knowledge of the behavior of other macroeconomic variables can be

helpful in determining the relationship between the oil prices and stock market activity.

In another study, Monjazeb and Hoseinzadeh (2013) investigated the impact of oil

revenue on the Tehran stock market dividend and price index. They used a VAR model to

estimate Tehran stock price index in its own lag, as well as oil income, global gold price, and

private sector investment. They developed their model as follows:

TEDPIX = C + β1TEDPIX(-1) + β2Gold Price + β3GDP + β4Oil + β5INVE + β6DUMGH + ϑ

They concluded that all variables examined in the study significantly impact the stock price

index. They claimed that oil revenue, GDP and gold price have positive impact, while

investment of private investors has a negative impact of the stock index.

Other Macroeconomics and the Stock Market

In contrast to the relationship between oil prices and the stock market, which has been at

the center of attention of numerous economics researchers throughout the world, the effect of

gold and real estate on the stock market has been much less studied. Only a few papers have

been published regarding these relationships, and the available papers are limited to the single

the country of study (mainly the United States), and cannot be generalized to other countries or
MACROECONOMICS AND THE STOCK MARKET 28

regions. That is because the outcome in each country can be different, depending on the

structure of the stock, gold, and real estate market within each country.

Gilmore, McManus, Sharma, and Tezel (2009) examined the relationship between the

global market gold price and the gold mining stock price and their impact on the stock market in

general. They used morning and afternoon price of gold on the London market, the stock price

of GOX and HUI as two gold mining companies, and three stock market indices representing

large, mid, and small capitalization companies to develop a multivariate cointegration model to

analyze the relationship between these variables. They also used the IRF and VDC techniques to

measure the nature and severity of the reaction of each variable to the shock to the other

variables. Running the model based on their data, they found evidence of cointegration among

all variables. They also found that the price of gold and the stock price of large cap companies

adjust to disturbances to restore the long term relationship between the variables.

Rad (2011) studied the impact of macroeconomic variables on the stock market in Iran

using the monthly data from 2001 to 2007. Rad used CPI, monetary base liquidity (M2), and

foreign exchange rate as macroeconomic variables to measure their impact on TSE. He

concluded that the response of the stock market price index to a shock to the selected variables

was weak and took four months to die out. He also concluded that the share of the selected

macroeconomic variables on the fluctuation of the stock market was about 12%, and these

macroeconomic variables do not play an important role in the fluctuation of TSE prices.

Shafiee and Tpal (2010) investigated the global gold market and gold price forecasting

based on the data from 1968 to 2008. By analyzing supply and demand of the gold market, they

showed that about 160,000 tons of gold has been mined from ancient times until the end of 2008,

of which 100,000 tons has been used by jewelry market, 30,000 tons used for industrial
MACROECONOMICS AND THE STOCK MARKET 29

purposes, and another 30,000 tons used by central banks as a reserve. Shafiee and Tpal

identified two reasons why short term increases in gold prices occur. The first reason is that

when the global financial market crashes and the global economy is in recession, then investors

tend to shift their money towards investments that are perceived to be very stable, predictable,

and to have little or no liability. Because gold is perceived as a commodity that never loses it

values, and possesses great stability, demand and investment in gold tends to rise when other

markets collapse. The second reason is the devaluation of the U.S. dollar, as well as

international inflation, which together tend to force large companies to purchase gold as a hedge

against fluctuation.

Shafiee and Tpal (2010) have also identified three reasons for long term increases in the

price of gold. These three reasons are: (a) reduction in the output of gold production from mines

due to high costs, decreased exploration, and difficulties in finding new deposits; (b) the keeping

of gold in portfolios by institutional and retail investing companies as a diversifying tool; and (c)

the relative ease of investment in gold by using exchange traded funds compared to other

financial markets. The same authors have also identified the price of oil and inflation rate as the

two main macroeconomic variables which influence the gold market. Their data and analysis

show a positive correlation between gold price and oil price () and a negative correlation

between gold price and cumulative inflation () over the previous four decades. However, the

correlation was not statistically significant. They calculated the ratio of oil price to one ounce of

gold as 11 for 1968 and 2008, which means that the price of these two commodities were

relatively stable for a long period of time.

Mishra, Das, and Mishra (2010) studied the relationship between gold prices and stock

market returns in India using the monthly data for the period of 1991 to 2009. They concluded
MACROECONOMICS AND THE STOCK MARKET 30

that there exists long run equilibrium between gold price and stock market returns in India. They

also concluded that gold price and the stock market in India are involved in a mutual relationship

in which both contain significant information for the prediction of the price of the other.

Raraga and Muharam (2013) investigated the impact of oil and gold prices on the stock

market index in Indonesia, based on the data for the period of 2000 to 2013. They developed a

vector error correction model using Jakarta Composite Index, Indonesian currency to the U.S.

dollar exchange rate, crude oil price, and global gold price as variables. They also obtained the

same result as Mishra et al. (2010), reporting that there is a cointegration between oil price, gold

price, exchange rate, and stock market in Indonesia. By this they mean that over the long term,

those variables move through an equilibrium point. Raraga and Muharam conclude that neither

the oil price nor the world gold price has any significant impact on the stock market in Indonesia.

However, they found that the exchange rate has a significant impact on the stock market and vice

versa, and claim there is a causal relationship between the two.

Smith (2001) examined the price of gold and the stock market indices for the United

States for the period of January 1991 to October 2001. In order to analyze the behavior of the

gold market and compare it to the stock market during the crisis period, Smith reviewed the

reaction of the stock market of several countries, as well as reaction of the gold market

immediately before and during a week after the September 11, 2001 catastrophe. As further

support for previous studies, he noticed that at the time of market crisis when the stock market

moves down, the gold market moves upward. Smith then used the U.S. morning and afternoon

gold price, as well as the London and New York gold market, as four different indicators for the

price of gold, and six groups of stock market indices in the United States, such as S&P 500,

NASDAQ, NYSE, AMEX, RUSSEL 3000, and WILSHIRE 5000 Equity Exchange. Based upon
MACROECONOMICS AND THE STOCK MARKET 31

these variables, he developed a linear regression between the price of gold and the stock market

in the United States. Smith concluded that the short-term correlation between gold price and the

stock market returns is small and negative, and he reported no long term equilibrium between the

two variables.

Baur and Lucey (2010) investigated the behavior of the gold market and asked whether

gold behaves as a hedge (a security that is uncorrelated with stock or bonds), or a safe haven (a

security that is uncorrelated with stocks and bonds during a market crash). They used data from

the United States, the United Kingdom, and Germany to investigate the behavior of gold and its

relationship with the stock and bond market. They found that gold is a safe haven for stocks in

all three countries, it is a hedge in the United States and the United Kingdom, but it is not a safe

haven or hedge tool for bonds in any market. They also concluded that gold behaves as a safe

haven for a limited time (for approximately 15 days), but that in the long run, gold is not a safe

haven. They concluded that investors who keep the gold for longer than 15 days after the shock

would lose money on their gold investment. The real estate market displays a different structural

behavior compared to the stock market. This is because the real estate market has a long tail,

which means it takes long time for the market to correct itself by reacting to its own fluctuations

and to other macroeconomic factors.

Yahyazadehfar and Babai (2012) used monthly data from 2001 to 2012 to study the

relationship between stock market price, gold price, housing price, and interest rate in Iran. They

used the VAR method and the IRF method to measure the long-term relationship. They also

used the Johansen-Juselius cointegration test to measure the short-term relationship among these

variables. They found a positive relationship between housing price and stock market price and

a negative relationship between gold price and interest rate with stock price.
MACROECONOMICS AND THE STOCK MARKET 32

Mehrara (2006) estimated the relationship between money supply, value of trade balance,

industrial production, and the stock market based upon the quarterly data from 1993 to 2004.

Based upon a long-run Granger non-causality test, Mehrara concluded that the Iranian stock

market is not informationally efficient. His based this conclusion on the observation that the

Iranian stock market is not a leading indicator for other economic variables. He also concluded

that for macroeconomic variables Granger-caused stock prices, no reverse causality was

observed.

There are several important facts about the structure, functions, and the relationship

between investment markets and microeconomics in Iran that need to be considered when

investigating the market behavior for the country. The hierarchy and structure of political power

and the relationship between state and society, as well as the patterns of economic and political

rules and regulations create unique environment in each country. Furthermore, the economic

factors and markets simply do not behave the same in Iran as they do in Western capitalist

countries.

One very important example of such a difference is that there is no option to purchase

gold stocks, real estate stocks, or foreign currencies on the Iranian stock market. This creates a

very different investment climate within Iran compared to most Western nations, in which

institutions such as exchange traded funds, mutual funds, REITs, and other tools make it simple

to invest in gold, real estate, or foreign currency through the stock market. In other words, gold,

real estate, and foreign currencies can be considered as a part of the stock market in most

countries with established advanced stock market, but in Iran, those markets are totally separate

from the stock market, and must therefore be considered as parallel, alternative, or substitute

investment markets to the stock market. If an investor decides to buy gold in Iran, he or she
MACROECONOMICS AND THE STOCK MARKET 33

needs to go to the gold market, pay cash and physically purchase the gold (in contrast to buying

gold mine shares by computers online, as would be done in Europe or the United States).

Likewise, the real estate market is totally different in Iran than in the United States and

most other Western countries. While in the United States, banks and other financial institutions

act as lenders, and provide a relatively high percentage of the cost of the home (in some cases up

to 90% of the price), with banks holding the title until the loan is eventually paid off many years

later. In contrast, in Iran, real estate transactions almost always occur as completely cash

purchases, which mean the buyers pays the whole price of the property in cash in one lump sum,

and receives the title immediately. As a result, other microeconomic factors, such as interest

rate, lender’s minimum, reserve requirements by Federal Reserve definitely impact the real estate

market in the United States. However, there is no such relationship between these

macroeconomic variables and the Iranian real estate market.

Based on the above facts and the best knowledge of the author, this present paper is

unique within the literature. This is because it considers oil dollar inflow, gold price, real estate

prices, and foreign exchange as several parallel investing markets, and uses all of these factors to

study the behavior of the Iranian stock market.

MACROECONOMICS AND THE STOCK MARKET 34

CHAPTER III

Research Model, Data, and Methodology

This study first examines the impact of oil revenue, followed by real estate, foreign

exchange rate, and GDP on the Iranian stock market, then ceteris paribus will substitute the oil

revenue with gold price and examine the results. In the final stage, the study runs the model for

all variables and compares the outcomes of each test in order to examine the possible impact of

the behavior of various combinations of markets on the stock market.

This study employs the autoregression ordinary least square approach, which is a very

common mathematical and econometric tool, to study the time series data and evaluate the

impact of economic variables on themselves and upon each other. At a later point in the study

VAR model will be run and the study will use different techniques, such as IRFs and VDC, to

measure the impact of the variables upon each other. The model is designed to estimate the

stock market reaction to different macroeconomic variables, as well as its reaction to its own lags

based on the data from Iranian economy.

An autoregression is a regression model in which Yt is regressed against its own lagged

values. In a first order autoregression AR(1), ϓt is regressed against ϓt–1. The population of

AR(1) model is designed as:

ϓt = β0 + β1ϓt-1 + Ԑt (3.1)

in which the AR(1) model can be estimated by OLS regression of ϓt against ϓt–1. When

analyzing macroeconomic data, it is important to make sure that the data are stationary, which

means their behavior is not influenced by any trend, seasonal factor, or cyclical impact. While

this test will be discussed in detail later in this chapter, it is important to consider that the data
MACROECONOMICS AND THE STOCK MARKET 35

which are used in this model are the first difference of the logarithm of the available data. The

first difference of the series ϓt is the difference between data ϓt and ϓt-1; that is:

Δϓt = ϓt – ϓt-1 (3.2)

From Equation 3.2, the first difference of the logarithm of ϓt can be developed as follows:

Δlnϓt = lnϓt – lnϓt-1 (3.3)

This study will develop four different models which will all be discussed in the following

pages.

Model 1: The Impact of Oil Revenue on the Stock Market

Based on this model, stock market activity is assumed to be a dependent variable, the

function of which depends on other macroeconomic independent variables such as the housing

market, GDP, CPI, oil revenue, and foreign exchange rate. The general model can be formulated

as in Equation 3.4 shown below:

𝑆𝑡𝑜𝑐𝑘(𝑡) = 𝛶0 + ∑𝑛𝑗=1 𝛶𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + 𝜉𝑡 (3.4)

where Stock(1t) represents the stock market price index in Iran at time t, ϓ0 is constant, ϓj is the

parameter which needs to be estimated, ξt is the error term, and macroecon is the group of

several macroeconomic variables identified and used to estimate the outcome. When developing

models to test the regression behavior of variables in econometric, one of the issues usually

discussed is to clarify whether variables are endogenous or exogenous. By definition,

endogenous variables are those which will be determined by the model, while exogenous

variables are those which are assumed to be determined by factors outside of the model. This

paper assumes stock market activity to be an endogenous variable, whose behavior will be

determined based on the outcome of the model. It is further assumed the rest of the variables are

exogenous, which means they assumed to be determined by factors external to the model. The
MACROECONOMICS AND THE STOCK MARKET 36

model then was developed by using the macroeconomic variables and formulated as Equation

3.5, as displayed below:

Stockt =ɑ0 + α1oilt-k + α2houset-k + α3exchanget-k+

α4gdpt-k + α5cpit-k + α6Interestt-k + ɛt (3.5)

where α0 is constant, α1 through α6 are parameters to be determined, k is the appropriate lag, ɛ is

the error term, and oil, house, interest, exchange, gdp, and cpi represent the logarithm of the first

difference of oil revenue, housing activity, long term interest rate, foreign exchange rate, GDPs,

and CPI of Iran, respectively.

Model 2: The Impact of Gold Price on the Stock Market

After running the model to test the impact of oil revenue on the stock market, this project

will test the impact of gold price on the stock market in Iran. To do so, the model keeps all

things equal and exchanges the oil revenue data with a gold price series. The purpose of this test

is to evaluate whether the gold price per se (without considering the impact of oil revenue) has

any impact on the stock market. The available data for this model was not stationary, and

therefore was designed as the first difference of the logarithm of available data, which is

displayed below as Equation 3.6. The general formula of the available data used to test the OLS

relationship between variables can be developed as follows:

𝑆𝑡𝑜𝑐𝑘(𝑡) = 𝜑0 + ∑𝑛𝑗=1 𝜑𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + 𝜔𝑡 (3.6)

where φ0 is constant, φj is the parameter which needs to be determined, ωt is the error term, and

macroecon is a group of macroeconomic variables such as gold price, GDP, CPI, foreign

exchange rate, and housing activities. By entering the available variables of the model, the

equation can be developed as follows:

MACROECONOMICS AND THE STOCK MARKET 37

Stockt = β0 + β1goldt-k + β2houset-k + β3exchanget-k + β4gdpt-k + β5cpit-k + β6interestt-k

+ ωt (3.7)

where β0 is constant, β1 through β6 are parameters to be determined, k is the proper lag, ω is the

error term, and gold, house, exchange, interest, gdp, and cpi are the first difference of the

logarithm of gold price, housing activity, foreign exchange rate, long term deposit interest rate,

GDPs, and CPI, respectively. As can be seen in the model, stock market is designed to be a

dependent variable to the other macroeconomic variables, and gold price has been designated as

the first independent variable followed by other macroeconomic variables. This has been done

to test whether there is any statistically significant relationship between the gold market and

other variables upon the activities of the Iranian stock market.

Model 3: The Impact of Oil Revenue, Gold Price Fluctuation, and Other Macroeconomic

Variables in Stock Market

In the third stage of the test, this study runs all variables, including oil revenue and gold

price, together in the model. For the final model, as in the two previous models, this study first

uses the OLS model and will then run the VAR model to use the IRF and VDC tool to analyze

the relationship between the variables. The general model for the final stage is:
𝑛
𝑆𝑡𝑜𝑐𝑘(𝑡) = Ӯ0 + ∑𝑗=1 Ӯ𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + Ӫ𝑡 (3.8)

where Stock(t) is the stock market activity in Iran, Ӯ0 is constant, Ӯj is to be determined, Ӫt is the

error term and macroecon is the group of macroeconomic variables, which are identified in the

data description. After inserting the variables into the model, the final model can be developed

as:

Stockt= χ0 +χ1oilt-k +χ2goldt-k +χ3houset-k +χ4exchanget-k +χ5gdpt-k +χ6cpit-k +

χ7interestt-k + ςt (3.9)
MACROECONOMICS AND THE STOCK MARKET 38

where χ0 is constant, χ1 through χ7 are parameters to be estimated, ςt is the random error term, k

is the proper lag, and Stock, oil, gold, house, exchange, dgp, cpi, and interest represent the first

difference of the logarithm of the activity of the Iranian stock market, oil revenue, gold price,

housing activities, foreign exchange rate, GDPs, CPI and long term deposit interest rate,

respectively. This final model considers the stock market as being endogenous to the model and

dependent to other macroeconomic variables in the model. The VAR model will be run on the

Equation 3.9 and IRFs, and VDC tests will be based on the outcome of this equation.

Model 4: The Impact of the Macroeconomic Variables on the Stock Market Volatility in

Iran

The fourth and final model is designed to examine and measure the relationship between

the selected macroeconomic variables and the stock market volatility. The standard deviation of

the stock market, based on the annual data, was calculated and then transformed into the

quarterly data series as a proxy of market volatility. In this case, the market volatility, not the

stock price index, has been examined against the other variables. The purpose of this model is to

measure the rate of the stock price variation as it relates to the behavior of selected

macroeconomic variables. The model is designed as follows:

Volatilityt= Ө0 +Ө1oilt-k +Ө2goldt-k +Ө3houset-k +Ө4exchanget-k +Ө5gdpt-k +Ө6cpit-k +

Ө7interestt-k + Ωt (3.10)

where Ө0 is constant, Ө1 through Ө7 are coefficients to be determined. The terms oil, gold,

house, gdp, cpi, and interest refer to Iranian oil revenue, gold price, housing activities, foreign

exchange rate, GDPs, CPI, and long-term deposit interest rate respectively, and Ωt is the error

term.
MACROECONOMICS AND THE STOCK MARKET 39

Methodology

VAR method has been widely used by researchers to study the relationship between

variables of time series data. The method was developed by Sims (1980), who believed that the

model can provide a coherent and credible approach to data description, forecasting, structural

inference, and policy analysis. The model is designed to analyze the regression of each variable

upon itself and upon other variables, which means each variable in the model is a linear function

of past lags of itself and past lags of the other variables. For example, the VAR model of order 1

of the two different time series as χt,1 and χt,2 that is VAR(1), can be developed as follows:

χt,1 = α1 + Ѳ11χt-1,1 + Ѳ12χt-1,2 + ὥt,1 (3.11)

χt,2 = α2 + Ѳ21χt-1,1 + Ѳ22χt-1,2 + ὥt,2 (3.12)

where each variable is a linear function of the lag 1 values for all variables in the set. By

combining the two equations, the general VAR model can be finalized as:

Ѵτ = λφτ + ϕѴτ-1 + ωτ (3.13)

where φτ = (1, τ)֜ represents the term that simultaneously fit the constant and the trend.

It is very important to consider that when using VAR, there are no dependent or independent

variables, but rather there are endogenous and exogenous variables. It is very difficult (and in

some cases impossible) to interpret the coefficient of VAR models, especially when there are too

many variables and too many lags involved. However, there are some strong and useful tools

which accompany the VAR system that can be used to investigate the relationship between the

variables. This study uses two of those powerful methods: IRFs and VDCs. IRF is the system

that measures the reaction of a variable to a one-time shock of one standard deviation of other

variables. For example, this tool can give us the idea of how the stock market reacts to a one

standard deviation shock separately to oil revenue, gold price, or housing activity. The other
MACROECONOMICS AND THE STOCK MARKET 40

useful tool, VDC can show us what percentage of the reaction of each variable in each lag is the

under influence of the activity of which other variable. For example, the tool can give us an

approximate estimate that what percent of stock market behavior is a response to oil revenue in

the first, second, third lag and so on. Both these tools have been widely used by researchers and

economists to analyze time series data in econometrics.

Data

This study was designed by using the quarterly observations of Iranian macroeconomic

variables from 1990 to 2012. The source of all the data in this study is the time series database

of the Central Bank of Iran except the stock index price which was found on the TSE website

(ew.tse.ir/en). Some data in the study was not available as quarterly records. For example, the

housing market data were available only as annual records. In order to convert the annual

records to quarterly observations, this study follows the quadratic match average approach,

which performs a proprietary local quadratic interpolation of the annual frequency data to fill in

the quarterly observations. Table A1 shows the raw data, and Table A2 contains the descriptive

statistics of all variables. The stock market activity indicator used in this study is the total share

price index of stock market in millions of rials. This paper used Iranian oil revenue instead of

global oil price. This was done because Iran oil exports were under the influence of global

sanctions during a portion of the study, so the oil cash inflow might fluctuate based on the

volume of export regardless of the global oil price. The data in the table is provided as billions

of rials per barrel. The housing activity indicator is the annual number of residential units

completed by the private sector in the urban areas in Iran during the period of study. The official

foreign exchange rate in Iran is controlled by Central Bank of Iran, and is not available to the

public for investment or hedge purposes. There are limitations even on exchanges for travel
MACROECONOMICS AND THE STOCK MARKET 41

purposes. Thus, there exist in Iran a strong active parallel black market for foreign currency

which is dominated by the U.S. dollar. Foreign currency is mainly used by currency day traders

and by ordinary local investors as a hedge against the depreciation of the rial. To measure

market activity, this study considers the unofficial black market exchange rate of the U.S. dollar

and one thousand Iranian rials. The gold price used within the study is the price of the old design

coin, which is commonly used for investment by local investors and is represented in the table by

1 million Iranian rials. This study uses the price of gold coins available in the Iranian market

instead of the global gold price. This is done to avoid the possibility of that there might be a

substantial difference between the price of gold inside and outside the country. CPI is the

official data from the Iran Central Bank, which sets the 2004 as the 100 bases (2004 = 100).

From the same source, an estimate of GDP is derived, based on billions of rials, and the year

1997 has been set as the 100 base year (1997 = 100). The nature of all data and its structure will

be discussed in detail in Chapter 4.

MACROECONOMICS AND THE STOCK MARKET 42

CHAPTER IV

Empirical Results

In this section, the data structure and descriptive statistics of the data will be analyzed,

and the time series properties of each variable will be checked for the possible existence of unit

root between variables. After this, the data will be entered into models, the models will be run,

and the outcome will be discussed. Finally, the VAR model will be performed, and IRFs and

VDCs outcomes will be discussed and analyzed.

Variance Inflation Factors

When analyzing time series data, in order to make sure the outcome is reliable and not

misleading, it is important to perform several tests before running the final model. One of the

risks when analyzing the time series is the possibility of the existence of collinearity between

variables, collinearity can increase estimates of parameter variance, and yield models which,

despite having a high R2, nevertheless yield results in which none of the variables are statistically

significant, and in which parameter have incorrect signs and implausible magnitude (Belsley et

al. 1980; Greene, 1993; O’Brien, 2007).

In a regression model, R2 represents the percentage of the variation within the data that

can be explained by the model outcome, and represents the proportion of the variance of an

independent variable that is associated with the other independent variables in the model. R2

measures the collinearity of the ίth independent variable with the other independent variables

within the model. In a regression analysis, tolerance is defined as one minus the proportion of

variance that each variable shares with other independent variables. This can be developed as:

Tolerance = 1 - 𝑅𝑖2 (4.1)

MACROECONOMICS AND THE STOCK MARKET 43

Equation 4.1 represents the proportion of variance in the ίth independent variable that is

not related or collinear to other variables. The variance inflation factor (VIF) is the reciprocal of

the tolerance:
1
VIF = (4.2)
1− 𝑅𝑖2

For example, if there is 90% collinearity between independent variables (extremely collinear),

the tolerance is .10 and the VIF is equal to 10. Likewise, a 75% collinearity leads to a tolerance

of .25, and the VIF would be 4. When interpreting the VIF, a VIF equal to 1 is usually

interpreted to mean there is no correlation between the variables, a VIF less than 5 is considered

to indicate a moderate degree of correlation, and a VIF between 5 and 10 is interpreted as

indicating a high degree of correlation between independent variables. To ensure the lack of

collinearity between the available variables, this study tested the data, and the results are

presented in Table 2.

Table 2

Variance Inflation Factor Test Result

OIL GOLD HOUSE EXCH CPI GDP Interest

VIF 1.69 1.38 1.57 1.04 1.36 1.23 1.18
Note. VIF = variance inflation factor; OIL = oil revenue; GOLD = gold price; HOUSE = housing activities; EXCH
= foreign exchange rate; CPI = consumer price index; GDP = gross domestic products; Interest = long-term interest
offered in Iran.

As can be seen, all VIFs are between 1 and 2. This indicates the assumption of noncollinearity is

not violated by any of the variables.

Unit Root Test of the Data

Another concern when analyzing time series data is the unit root situation between

variables. The theory of autoregressive equations is based on stationary time series. A time
MACROECONOMICS AND THE STOCK MARKET 44

series is stationary when its properties do not depend on the time at which the series is observed.

In other words, mean and autocovariance in a nonstationary series depend upon the time at which

data were observed. A common example of nonstationary data in finance and econometric is the

random walk equation:

γτ = γτ-1 + ȩτ (4.3)

where ȩ is white noise or a stationary random disturbance term. The forecast value of the series γ

is conditional on time τ. The existence of a trend and seasonality are examples of why unit root

condition happens within data. One method for converting the raw nonstationary data into

stationary data is to use the differences between moments to be able to fit them in the model.

The first difference is the change between one observation and the next. Expressed in

mathematical terms, this would be as follows:

ўτ = γτ – γτ-1 (4.4)

In some cases, the differenced data will not appear stationary, and it may be necessary to

calculate the second difference of data. For example:

Ӳτ = ўτ – ўτ-1 (4.5)

Using the second equation, we can formulate the second difference equation as follows:

Ӳτ = γτ - 2γτ-1 + γτ-2 (4.6)

Dickey and Fuller (1979, 1981) improved the mathematical test for determining whether

there is any unit root between the time series. Their improved method has been named after

these authors, and is therefore called augmented Dickey-Fuller. The results of the augmented

Dickey-Fuller test using the data for this study is provided in Table 3.
MACROECONOMICS AND THE STOCK MARKET 45

Table 3

Unit Root Tests of Variables

Levels: ADF Lag Length

Stock 1.230044 2
Oil_Rev. -2.36949 2
Gold 2.545231 2
Housing 0.6024 2
Exchange Rate 1.512676 2
CPI 43353206 2
GDP 1.326702 2
Interest 2.270317 2
First Differences:
Stock 1.89424 2
Oil_Rev. -4.2616*** 2
Gold 0.60075 2
Housing -3.0859** 2
Exchange Rate 4.80336 2
CPI -0.0449 2
GDP -4.575 2
Interest -2.8395** 2
Second Differences:
Stock -7.9715*** 2
Oil_Rev. -5.7322*** 2
Gold -9.27*** 2
Housing -11.811*** 2
Exchange Rate -4.6805*** 2
CPI -9.4937*** 2
GDP -7.4713*** 2
Interest -6.7413***
***p < .01 and **p < .05 were the
significance levels for the rejection of the null
hypothesis of unit root for the ADF.

OLS Model Results

When the time series properties of all variables were determined to be stationary, the

OLS regression model was applied to measure the impact of all variables on the stock market in

Iran. This method was used separately for all four models described in Chapter 3 (3.5, 3.7, 3.9,

and 3.10), and the results are discussed in the section “OLS Results of Model 1.” The VAR

model and tests will be discussed in the section “OLS Results of Model 2.”
MACROECONOMICS AND THE STOCK MARKET 46

OLS Results of Model 1: The Impact of Oil Revenue on the Stock Market in Iran

As with most other econometric and statistical methods, there are several assumptions

about OLS which need to be followed in order to obtain reliable results. By analyzing the

outcome, this study will assure that none of the required assumptions are violated. Table 4

shows the results of the OLS regression for Model 1 variables.

As explained previously, this study conducted several tests on the above data to make

sure none of the OLS assumptions were violated. First and foremost, the model does not seem

robust. This is because the R2 is only .30, which indicates the model and the outcome can

explain only 30% of the data and the relationship between the variables. To address the other

assumptions, this study then conducted the normality test. In this study, the assumption is that

the mean of all samples are distributed normally.

MACROECONOMICS AND THE STOCK MARKET 47

Table 4

Effects of Oil Revenue and Other Macroeconomics on the Stock Market: 𝑆𝑡𝑜𝑐𝑘(𝑡) = 𝛶0 +
∑𝑛𝑗=1 𝛶𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + 𝜉
Variable Coefficient Std. Error t Statistic Prob.
OIL 0.003632 0.004520 0.803543 0.4244
HOUS -0.210901 0.054600 -3.862652 0.0003
EXCH 0.009216 0.017246 0.534400 0.5948
CPI -0.056831 0.033024 -1.720900 0.0898
GDP -0.001192 0.001160 -1.028639 0.3077
Interest 0.042853 0.024928 1.719119 0.0901

R2 .303002
Adjusted R2 .252495
SE of regression .063403
Sum squared residuals .0277377
Log likelihood 103.5746
Durbin-Watson stat 0.917440
Note. Stock = Iranian Stock Market; OIL = oil revenue; HOUS = housing activity; EXCH =
foreign exchange rate (EXCH); CPI = consumer price index (CPI); GDP = gross domestic
products.
***p < .01.

One way to examine this assumption is to use a histogram and use the Jarque-Bera

indicator. This Jarque-Bera indicator is named after the two mathematicians who invented it,

and indicates the skewness and kurtosis of the distributed data. This Jarque-Bera indicator is

performed as a way to test for normality. Figure 2 provides the results of the histogram and

Jarque-Bera test.
MACROECONOMICS AND THE STOCK MARKET 48

30
Series: Residuals
Sample 1991Q3 2011Q4
25 Observations 75

20 Mean 0.006241
Median 0.000744
Maximum 0.326860
15 Minimum -0.234159
Std. Dev. 0.060900
Skewness 1.279471
10
Kurtosis 14.77016

5 Jarque-Bera 453.3904
Probability 0.000000
0
-0.2 -0.1 0.0 0.1 0.2 0.3
Figure 2. The histogram and Jarque-Bera indicator. Normality assumption test.

As can be seen, there are problems with the normality distribution of the data. This is

because we have an extremely large Jarque-Bera coefficient and kurtosis indicator. Another

assumption required for regression analysis is the heteroscedasticity of the data. The term

heteroskedastic is used to describe data in which the standard deviations of a variable are non-

constant during the specific period of time. Figure 3 shows an examples of heteroscedastic and

homoscedastic data distributions.

Homoscedasticity Heteroscedasticity
100 100

80 80

60 60

40 40

20 20

0 20 40 60 80 100 0 20 40 60 80 100

Figure 3. Examples of homoscedasticity versus heteroscedasticity distribution of data.

This study conducted the Breusch-Pagan-Gofrey heteroscedasticity test to examine the

data, and the result is displayed in Table 5.

MACROECONOMICS AND THE STOCK MARKET 49

Table 5

Heteroscedasticity Test: Breusch-Pagan-Godfrey

F statistic 2.101307 Prob. F(6,68) 0.0643

Obs*R2 11.73072 Prob. Chi-Square (6) 0.0683
Scaled explained SS 69.70211 Prob. Chi-Square (6) 0.0000

A probability of 6.8% indicates that the null hypothesis of homoscedasticity can be

rejected on 10% level of confidence, and the data set is heteroscedastic on that level, which is a

violation of the data distribution assumptions. In order to correct for these issues as part of the

process of developing the model, this study adds a new set of data “AR (1)” based on the

available variables. Auto regression (AR) data specifies that the output variable depends linearly

on its own previous values on a random term. Using that technique, the outcome is shown on

Table 6. This table summarizes the results of the OLS regression to measure the relationship

between oil revenue, housing activities, exchange rate, CPI, GDP, and Interest rate after the

required corrections have been made.

As can be seen, the OLS result demonstrates that oil, GDP, and housing activity are

statistically significant, and that GDP and housing activities have a negative impact while oil

price has a positive impact on the stock market price index. Housing activity has a stronger

coefficient (-.22), and has a severely and negative impact stock prices. On the other hand, the

positive sign and the coefficient of oil revenue and GDP can be interpreted as a direct

relationship between these two variables and the stock price index. Because it can explain about

88% of the relationship between the variables, the model seems to be robust.
MACROECONOMICS AND THE STOCK MARKET 50

Table 6

Effects of Oil Revenue and Other Macroeconomics on the Stock Market (adjusted)
𝑛

𝑆𝑡𝑜𝑐𝑘(𝑡) = 𝛶0 + ∑ 𝛶𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + 𝜉
𝑗=1
Variable Coefficient Std. Error t Statistic Prob.
OIL 0.01431*** 0.001762 8.119737 0.0000
HOUS -0.225515*** 0.01882 -11.98277 0.0000
EXCH 0.001804 0.004876 0.370069 0.7126
CPI -0.000374 0.01222 -0.030589 0.9757
GDP -0.001167*** 0.00035 -3.330977 0.0015
Interest 0.000650 0.010211 0.063678 0.9494
AR(1) 0.26959 0.037475 7.193906 0.0000

R-squared 0.875091
Adjusted R-squared 0.862609
SE of regression 0.020849
Sum squared residuals 0.026515
Log likelihood 167.3946
Durbin-Watson stat 2.308476

Note. Stock = Iranian Stock Market; OIL = oil revenue; HOUS = housing activity; EXCH =
foreign exchange rate; CPI = consumer price index; GDP = gross domestic products; Interest =
interest rate.
***p < .01.

In econometric data analysis, there are several tools and indicators that allow researchers

to correct their models and make sure that the model has not violated any of the econometric

assumptions. One important criterion that needs to be followed is the Durbin-Watson statistic

test, which is used to detect any autocorrelation or serial correlation between the data. It is used

to detect the correlation between values of the process at different times, as a function of the two

times or of the time lag. In other words, it detects the similarities between observations as the

function of the time lag between them. Generally, if the Durbin-Watson test statistic has a result

of less than 2, this is interpreted as evidence of a positive serial correlation within the data. The

results are presented later in this chapter. Because a Durbin-Watson result of 2.3 was obtained,
MACROECONOMICS AND THE STOCK MARKET 51

the clearly indicates that the model has not violated the test, and there is no evidence of any

positive serial correlation between the variables.

Another criterion used in statistical analysis is the Jarque-Bera test, which is also known

as “goodness of fit” test. As mentioned previously, the Jarque-Bera test measures whether the

data contains skewness and kurtosis or approximates a normal distribution. Figure 4 shows the

result of Jarque-Bera test on the data after adding the new AR(1) set of data.

14
Series: Residuals
12 Sample 1991Q4 2011Q4
Observations 67
10
Mean 0.001021
Median 0.000466
8 Maximum 0.048279
Minimum -0.048787
6 Std. Dev. 0.020017
Skewness 0.081665
4 Kurtosis 3.317837

Jarque-Bera 0.356487
2
Probability 0.836739
0
-0.04 -0.02 0.00 0.02 0.04
Figure 4. Histogram and Jarque-Bera test results (adjusted).

The probability of approximately 84% means that the null hypothesis of normality (that

the data distribution is normal) cannot be rejected.

Similar to the Durbin-Watson test, the Breusch-Godfrey test is another criterion which

considers the validity of some of the modeling assumptions in regression analysis, particularly

serial correlation within data. The result of the Breusch-Godfrey LM test is summarized in Table

7.
MACROECONOMICS AND THE STOCK MARKET 52

Table 7

Breusch-Godfrey LM Test Results

Breusch-Godfrey Serial Correlation LM

Test
F statistic 1.316745 Prob. F(4, 57) 0.2750

Obs*R-squared 5.598238 Prob. Chi-Square (4) 0.2312

Note. The probability of 23% indicates that the null hypothesis cannot be rejected.

As displayed on the table, the probability of approximately 23% of four lags of the data

clearly indicates that the null hypothesis of LM test cannot be rejected, and the model has not

violated this assumption.

Another concern in autoregression analysis is the heteroscedasticity of variables. This

refers to existence of multiple subpopulations which have different variabilities (such as variance

or any other statistical dispersion) from each other. The presence of such patterns can invalidate

the result of other statistical tests. For example, the result of OLS will be inaccurate because the

true variance and covariance are underestimated. There are several tests available to detect

heteroscedasticity. This study used the Breusch-Pagan-Godfrey test, and the result is shown on

Table 8.

Table 8

Heteroscedasticity Test Results

Heteroscedasticity Test: Breusch-Pagan-
Goddfrey
F statistic 0.961938 Prob. F(6, 60) 0.4587
Obs*R-squared 5.879423 Prob. Chi-square (6) 0.4368
Scaled explained SS 5.499755 Prob. Chi-square (6) 0.4815
Note. The null hypothesis of homoscedasticity with a probability of about 44% cannot be
rejected.
MACROECONOMICS AND THE STOCK MARKET 53

OLS Results on Model 2: The Impact of Gold Price on the Stock Market Activity in Iran

In this model, ceteris paribus, oil price was replaced by gold price to investigate whether

the gold price fluctuation has any impact on the stock market activity in Iran. As discussed

previously, the intention behind developing this model was to test the reaction of the two

investment markets (gold and the stock market) on the behavior of each other. The original data

was run in the model and the result is shown on Table 9.

Table 9

The Impact of Gold Price and Other Macroeconomics on the Stock Market Activity in Iran
𝑆𝑡𝑜𝑐𝑘(𝑡) = 𝜑0 + ∑𝑛𝑗=1 𝜑𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + 𝜔𝑡
Variable Coefficient Std. Error t Statistic Prob.

GOLD -0.005789 0.083045 -0.069711 0.9446

HOUSING -0.089556 0.052556 -1.704005 0.093
GDP -8.72E-05 0.001102 -0.079116 0.9372
EXCHANGE 0.009748 0.015253 0.639053 0.525
INTEREST 0.041118 0.021791 1.886907 0.0635
CPI -0.079523 0.029252 -2.718545 0.0083

R-squared 0.162524
Adjusted R-squared 0.100026
SE of regression 0.056249
Sum squared resid 0.211988
Log likelihood 109.6391
Durbin-Watson stat 0.768361

Note. Stock = Iran Stock Market; Gold = gold price; HOUS = Housing Activity; EXCH =
foreign exchange rate; CPI = consumer price index; GDP = gross domestic products; Interest =
interest rate.

As this model also shows, most of the OLS assumptions are violated. The Durbin-

Watson indicator of less than 1, with the R2 of 16.3% indicates that the results are not reliable.

Figure 5 presents the histogram and the Jarque-Bera indicator, which is further evidence that the

normality assumption has been violated.

MACROECONOMICS AND THE STOCK MARKET 54

28
Series: Residuals
24 Sample 1991Q3 2011Q4
Observations 73
20
Mean 0.006495
Median 0.000944
16 Maximum 0.280007
Minimum -0.224368
12 Std. Dev. 0.053866
Skewness 0.890185
8 Kurtosis 14.84647

Jarque-Bera 436.5054
4
Probability 0.000000
0
-0.2 -0.1 0.0 0.1 0.2 0.3
Figure 5. Histogram and Jarque-Bera test results.

To correct for the above-mentioned issues, this study once more adds a set of AR(1) data.

As discussed previously, autoregression is a model in which Ƴt is regressed against its own

lagged value. The number of lags in the model is called the “order” of the regression. AR(1)

represents the first order regression, which means the Ƴt is regressed against Ƴt-1. In most

cases, adding the AR(1) series of data to the model will increase the accuracy by improving

Durbin-Watson coefficient. This study also adjusted the data to account for some outliers, which

could potentially cause the normality distribution be violated. After adding the new set of AR

data and adjusting for outliers, the data was run again. The outcome is presented in Table 10.

As shown in the table, using the AR(1) technique solved all of the above-mention

problems. The Durbin-Watson indicator is now greater than 2, the R2 is relatively high. The

normality test and histogram are presented in Figure 6 with a small Jarque-Bera indicator (.88)

and high probability (64%). This indicates that the assumption of normality is not violated.
MACROECONOMICS AND THE STOCK MARKET 55

Table 10

The Impact of Gold Price and Other Macroeconomics on the Stock Market Activity in Iran
𝑆𝑡𝑜𝑐𝑘(𝑡) = 𝜑0 + ∑𝑛𝑗=1 𝜑𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + 𝜔𝑡. (Adjusted)

Variable Coefficient Std. Error t Statistic Prob.

GOLD -0.024621 0.03313 -0.743171 0.4604
HOUSING -0.186367 0.022375 -8.329107 0.0000
GDP -0.001021 0.000391 -2.61342 0.0114
EXCHANGE 0.001344 0.005055 0.265982 0.7912
INTEREST 8.88E-05 0.010027 0.00886 0.993
CPI -0.000823 0.012772 -0.064455 0.9488
AR(1) 0.233929 0.044829 5.21828 0.0000

R2 0.62277
Adjusted R2 0.583062
SE of regression 0.020709
Sum squared resid 0.024444
Log likelihood 161.0356
Durbin-Watson stat 2.502855
Note. Stock = Iran Stock Market; Gold = gold price; HOUS = housing activity; EXCH = foreign
exchange rate; CPI = consumer price index; GDP = gross domestic products; Interest = interest
rate.

16
Series: Residuals
14 Sample 1991Q4 2011Q4
Observations 64
12
Mean 0.001430
10 Median 0.000329
Maximum 0.047751
8 Minimum -0.052470
Std. Dev. 0.019645
6 Skewness -0.016548
Kurtosis 3.576100
4
Jarque-Bera 0.887965
2 Probability 0.641477
0
-0.04 -0.02 0.00 0.02 0.04
Figure 6. Histogram and Jarque-Bera test results (adjusted).

The heteroscedasticity test on Table 11 also indicates a probability of approximately

23%. This means the hypothesis of homoscedasticity cannot be rejected and the data are

homoscedastic.
MACROECONOMICS AND THE STOCK MARKET 56

Table 11

Heteroscedasticity Test Results on Model 2 (Adjusted)

Heteroscedasticity Test: Breusch-Pagan-Godfrey

F statistic 1.386332 Prob. F(6, 57) 0.2359

Obs*R2 8.15015 Prob. Chi-square (6) 0.2273
Scaled explained SS 8.291396 Prob. Chi-square (6) 0.2175
Note. The null hypothesis of homoscedasticity with a probability of about 23% cannot be
rejected.

The data display in Table 11 indicates that among all variables, only housing activity and

GDP have any significant impact of the stock market activity in Iran. Both of these variables

have a negative coefficient. The coefficient of the housing activity is greater than the coefficient

of GDP. This pattern can be interpreted to mean that when housing activity is slower, the stock

market activity is greater. The outcome also indicates that gold price fluctuation has no

significant impact on stock market activity. This may lead us to the conclusion that the

purchasing of gold as an investment tool is not a substitute for the stock market.

OLS Results on Model 3: The Impact of Oil Revenue, Gold Price Fluctuation, and Other

Macroeconomics on the Stock Market activity in Iran

The outcome of the OLS regression on the first two models indicated that oil revenue,

(but not gold), has a significant impact on the stock market activity. Housing activity and GDP

also were significant in the outcome of both models. The third model considers all the variables,

including oil revenue, gold price fluctuation, housing activity, and foreign exchange rate as most

common investment tools as well as GDP, interest rate, and CPI. This will measure if the

outcome would be different after considering all variables in one equation. The concept behind

the model is that since Iran is an oil-exporting country, and the Iranian stock market is a

relatively new investment tool, the relationship between the various named variables may be
MACROECONOMICS AND THE STOCK MARKET 57

different in Iran than the patterns reported in the oil-importing Western countries. This study

focuses on this third model since it involves all the variables.

In the next section, this study will measure the relationship between these variables and

stock market volatility. After this, the VAR model will be used in Model 3 to determine the

reactions of the stock market to a one standard deviation shock to other macroeconomic

variables.

As in previous sections, when the data was run in the model, the outcome indicated that

some OLS assumptions are violated. The result of the OLS outcome is provided in Table 12.

This information indicates that the model is not robust. The R2 result was only 31%, and a

Durbin Watson statistic of 0.91 indicates the data are not reliable.

Table 12

The Impact of Oil Revenue, Gold Price Fluctuation, and Other Macroeconomics on the
𝑛
Stock Market. 𝑆𝑡𝑜𝑐𝑘(𝑡) = Ӯ0 + ∑𝑗=1 Ӯ𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + Ӫ𝑡

Variable Coefficient Std. Error t Statistic Prob.

OIL 0.002911 0.004575 0.63632 0.5267
GOLD 0.091046 0.089901 1.012736 0.3148
HOUSING -0.209984 0.054597 -3.846032 0.0003
GDP -0.000809 0.001219 -0.663053 0.5095
EXCHANGE 0.010221 0.017271 0.591784 0.556
CPI -0.058908 0.033081 -1.780699 0.0794
INTEREST 0.043122 0.024924 1.730122 0.0881

R2 0.313359
Adjusted R2 0.252773
SE of regression 0.063391
Sum squared resid 0.273256
Log likelihood 104.1359
Durbin-Watson stat 0.915591
Note. Stock = Iranian Stock Market; OIL = oil revenue; Gold = gold price; HOUS = housing
activity; EXCH = foreign exchange rate; CPI = consumer price index; GDP = gross domestic
products; Interest = interest rate.
MACROECONOMICS AND THE STOCK MARKET 58

As part of the process of evaluating the reliability of data in this study, the outcome of

Model 3 also was tested for normality assumption. The histogram presented in Figure 7

indicates that the data are not distributed normally, the Jarque-Bera statistic is not sufficient, and

there are some outliers within the data.

25
Series: Residuals
Sample 1991Q3 2011Q4
20 Observations 75

Mean 0.006692
15 Median 0.000870
Maximum 0.322294
Minimum -0.239570
Std. Dev. 0.060392
10
Skewness 1.106247
Kurtosis 14.84516
5
Jarque-Bera 453.7592
Probability 0.000000
0
-0.2 -0.1 0.0 0.1 0.2 0.3

Figure 7. Histogram and normality test of the Model 3.

As in previous sections, for this set of data, this study follows the AR(1) technique to

correct for the violations of the assumptions about the normality of the distributions. As

discussed earlier in this chapter, the AR(1) was added to the model and the data was run. This

was done so that the model would estimate the outcome based not only upon the previous lags of

each model on each other, but also upon its own previous lag as well. The final outcome is

presented in the Table 13.

MACROECONOMICS AND THE STOCK MARKET 59

Table 13

The Impact of Oil Revenue, Gold Price Fluctuation, and Other Macroeconomics on the Stock
𝑛
Market. 𝑆𝑡𝑜𝑐𝑘(𝑡) = Ӯ0 + ∑𝑗=1 Ӯ𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + Ӫ𝑡

Variable Coefficient Std. Error t Statistic Prob.

OIL 0.014601 0.001873 7.797116 0.0000

GOLD -0.017613 0.03323 -0.530027 0.5981
HOUSING -0.225194 0.019558 -11.51431 0.0000
GDP -0.001231 0.000376 -3.27487 0.0018
EXCHANGE 0.001743 0.005049 0.345288 0.7311
CPI 0.000224 0.012732 0.017611 0.9860
INTEREST 0.000172 0.010283 0.016744 0.9867
AR(1) 0.264986 0.038959 6.801597 0.0000
R-squared 0.875692
Adjusted R-squared 0.860943
SE of regression 0.021148
Sum squared resid 0.026388
Log likelihood 167.5561
Durbin-Watson stat 2.316481
Note. Stock = Iranian Stock Market; OIL = oil revenue; Gold = gold price; HOUS = housing
activity; EXCH = foreign exchange rate; CPI = consumer price index; GDP = gross domestic
products; Interest = interest rate.

After make some adjustments for some outliers, in order to make sure that the data are

distributed normally, this study tested the data after adjustments. The results appear in the

histogram shown in Figure 8. The Jarque-Bera statistic and its probability are calculated to be

0.63% and 73%, respectively. It seems that the data is distributed normally, and the assumption

of normality is not violated.

Running the Breusch-Pagan-Godfrey heteroscedasticity test on the Model 3 data also

indicates that the data are homoscedastic (see Table 14).

MACROECONOMICS AND THE STOCK MARKET 60

16
Series: Residuals
14 Sample 1991Q4 2011Q4
Observations 67
12
Mean 0.000935
10 Median -0.000314
Maximum 0.049520
8 Minimum -0.048126
Std. Dev. 0.019973
6 Skewness 0.150826
Kurtosis 3.369999
4
Jarque-Bera 0.636200
2 Probability 0.727530
0
-0.04 -0.02 0.00 0.02 0.04

Figure 8. Histogram and normality test of the Model 3 (adjusted).

Table 14

Heteroscedasticity Test Results on Model 3 (Adjusted)

Heteroscedasticity Test: Breusch-Pagan-Godfrey

F statistic 1.310144 Prob. F(7, 59) 0.2616

Obs*R-squared 9.013475 Prob. Chi-square (7) 0.2517
Scaled explained SS 8.375771 Prob. Chi-square (7) 0.3006
Note. The null hypothesis of homoscedasticity with a probability of about 26% cannot be
rejected.

As shown in Table 14, the probability of the null hypothesis is 25% which means the

homoscedasticity of data cannot be rejected and data are not heteroscedastic.

Based on the data shown in Table 13, the only three variables that significantly impact

the stock market activity are oil revenue, housing market activity, and GDP. This data indicates

that oil revenue positively impacted the stock market, but the other two variables negatively

impacted the market. Of all these variables, the strongest coefficient belongs to housing market

activity

(-23%), while the GDP coefficient is very close to zero. The GDP has the least impact on the
MACROECONOMICS AND THE STOCK MARKET 61

market (-0.1%), and when considered in combination with other forces on the market, its impact

can be disregarded.

The gold price coefficient is -1.8%. However, the probability of the null hypothesis is

about 60%, which means the null hypothesis cannot be rejected, and that the price of gold does

not have a significant impact on stock market activities. This finding can lead us to the

conclusion that gold is not being used as a substitute investment for the stock market. This

conclusion is robust, for the result was the same for the other two models in which oil revenue

was also not considered as a variable. In other words, regardless of the oil money inflow to the

Iranian economy, investors in the Iranian stock market do not chose gold as a substitute to the

stock market.

OLS Results on Model 4: The Impact of Oil Revenue, Gold Price Fluctuation, and Other

Macroeconomics on the Stock Market Volatility in Iran

This final model was developed in the response to the question of whether any of the

variables in the study have significant impact on the volatility of the stock market. Volatility is

defined as the measurement of the change of the market price during a defined period of time for

a given set of return. To estimate stock market volatility, this study calculated the percentage

change of the annual stock index price, then calculated the standard deviation of consecutive

years. The standard deviation of the return was then transformed into quarterly data using the

quadratic average of the observed data. This quarterly standard deviation of the return was used

to measure the volatility of the market against the activity of the other variables. As with

previous models, the result of the raw data appears in the Table 15. This indicates that all of the

previously named assumptions are violated, the raw data need to be adjusted for outliers, and the

model needs to be run based on the variable’s own previous lag behavior.
MACROECONOMICS AND THE STOCK MARKET 62

Table 15

The Impact of Oil Revenue, Gold Price Fluctuation, and Other Macroeconomics on the Volatility
𝑛
of the Stock Market in Iran: 𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦(𝑡) = Ӯ0 + ∑𝑗=1 Ӯ𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + Ӫ𝑡
Variable Coefficient Std. Error t Statistic Prob.

OIL -0.045267 0.003674 -12.3216 0.0000

GOLD -0.111327 0.054242 -2.052399 0.0444
HOUSING 0.138421 0.035639 3.88394 0.0003
GDP -0.004851 0.000757 -6.411939 0.0000
INTEREST 0.005095 0.014484 0.351759 0.7262
EXCHANGE -6.36E-05 0.009844 -0.006463 0.9949
CPI 0.062509 0.01969 3.174714 0.0023
R2 0.847236
Adjusted R2 0.832453
SE of regression 0.03595
Sum squared resid 0.080128
Log likelihood 135.2522
Durbin-Watson stat 1.256909
Note. The Durbin-Watson statistic is not significant and the outcome is not reliable.

The histogram in the Figure 9 also shows that the Jarque-Bera test and its probability is

not sufficient, and that the data deviate slightly from a normal distribution. These results

indicate the existence of some outliers within the data, and that the data need to be adjusted for

distribution. As discussed earlier, adding a set of data as AR(1) to the model would allow the

model to estimate the outcome based on the regression of previous lag of all other variables, as

well as previous lag of its own variable.

 MACROECONOMICS AND THE STOCK MARKET 63

20
Series: Residuals
Sample 1991Q2 2011Q4
16 Observations 69

Mean 0.004989
12 Median 0.000631
Maximum 0.117965
Minimum -0.070145
Std. Dev. 0.033957
8
Skewness 1.232838
Kurtosis 5.862449
4
Jarque-Bera 41.03538
Probability 0.000000
0
-0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12

Figure 9. The histogram and Jarque-Bera result of the Model 4.

After adjustments for outliers and for distribution of the data, the test was run again, and

the results are presented in Table 16.

Table 16

The Impact of Oil Revenue, Gold Price Fluctuation, and Other Macroeconomics on the Volatility
𝑛
of the Stock Market in Iran: 𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦(𝑡) = Ӯ0 + ∑𝑗=1 Ӯ𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + Ӫ𝑡 (Adjusted)
Variable Coefficient Std. Error t Statistic Prob.

OIL -0.039892 0.002451 -16.27446 0.0000

GOLD -0.093921 0.029373 -3.197534 0.0024
HOUSING 0.160308 0.023457 6.834009 0.0000
GDP 0.00028 0.000569 0.490892 0.6257
INTEREST -0.005137 0.009661 -0.531714 0.5973
EXCHANGE 0.005217 0.004258 1.225287 0.2262
CPI -0.036148 0.016293 -2.218629 0.0311
AR(1) 0.363881 0.124236 2.928951 0.0051
R-squared 0.941227
Adjusted R-squared 0.932999
S.E. of regression 0.018255
Sum squared resid 0.016663
Log likelihood 154.1973
Durbin-Watson stat 1.90619

Note. The Durbin-Watson statistic is improved and R2 is close to 2.0.

MACROECONOMICS AND THE STOCK MARKET 64

The histogram after adjustment also indicates that data are normally distributed. The

result is shown in Figure 10.

12
Series: Residuals
Sample 1991Q3 2010Q1
10 Observations 58

8 Mean 0.002520
Median 0.003100
Maximum 0.039470
6 Minimum -0.039294
Std. Dev. 0.016908
Skewness 0.080110
4
Kurtosis 3.235594

2 Jarque-Bera 0.196174
Probability 0.906570
0
-0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04

Figure 10. The histogram and Jarque-Bera result of Model 4.

Based on the above results, the outcome of the test is reliable. The outcome indicates that

oil revenue, gold price, housing activity, and CPI significantly impact the volatility of stock

market returns. The results demonstrate that oil revenue, gold price, and CPI have a negative

impact and housing activity has a positive impact on the stock return volatility. That means that

when oil revenue and gold price are lower, the stock market is more volatile. Furthermore, this

means that when housing activity is higher, the stock market is more volatile. The result of this

model, as well as all previous models, will be discussed at the end of this chapter.

VAR Results

This section presents the results of running the VAR on Model 3, which includes all of

the variables. In most of these cases, the VAR outcome is very difficult (if not impossible) to

interpret, and therefore can be very confusing. In other words, the VAR coefficient does not

capture the full effect of an independent variable. However, there are several strong tools in the
MACROECONOMICS AND THE STOCK MARKET 65

VAR model which are frequently used to capture the relationship among the behaviors of the

variables within the model.

This study approached two of the models in the VAR system. In this section, the study

first ran the VAR model and used the IRFs of the variables. IRFs are the reaction of each

variable to a 1 standard deviation shock to another variable. With this tool, we can separately

measure the reaction of the stock market to a one standard deviation shock to all other variables.

Next, this study uses VDC analysis, which allows us to measure what percentage of the

fluctuation of each variable comes from another variable during different time periods. For

example, in this case, we can separately measure what percent of the fluctuation of the stock

market price index is due to the shock to oil revenue, gold, and other variables. The result of

each test will be discussed at the end of this section. In order to run the VAR model, the

researcher first used the lag length criteria technique to determine how many lags are necessary

to use with the model to obtain the best and most reliable outcome. Table 17 presents the log

length selection of the data based on various criterion.

Table 17

Lag Length Selection Criteria

Lag LogL LR FPE AIC SC HQ

0 -61.97637 — 5.49E-09 3.682967 4.027722 3.805628
1 66.60995 196.2633 1.97E-10 0.283687 3.386481 1.387637
2 152.9847 95.46679 9.74E-11 -0.89393 4.966905 1.19131
3 305.7599 104.5304* 4.31e-12* -5.566312* 3.052563* -2.499782*
Note. LR = sequential modified LR test statistic (each test at 5% level); FPE = final prediction
error; AIC = Akaike information criterion; SC = Schwarz information criterion; HQ = Hannan-
Quinn information criterion.
MACROECONOMICS AND THE STOCK MARKET 66

As can be seen, almost all criteria recommend the third lag as the best selection for the

VAR model. After adjusting for the lag length, the VAR model was run and the result is

presented in Table 18.

Table 18

VAR Result

Vector Auto-Regression Estimates

STOCK OIL GOLD HOUSING GDP INTEREST EXCHANGE CPI
STOCK(-1) 0.126812 4.080718 0.090081 -1.432942 2.670184 -2.585576 -4.91276 1.625579
-0.59626 -10.1314 -0.62694 -2.22466 -14.3726 -4.38778 -7.73264 -1.53467
[ 0.21268] [ 0.40278] [ 0.14368] [-0.64412] [ 0.18578] [-0.58927] [-0.63533] [ 1.05924]
STOCK(-2) -0.053755 -5.964615 0.442821 -0.130761 -3.058652 0.83676 2.963743 2.082665
-0.63667 -10.8179 -0.66943 -2.37542 -15.3466 -4.68513 -8.25667 -1.63867
[-0.08443] [-0.55136] [ 0.66149] [-0.05505] [-0.19931] [ 0.17860] [ 0.35895] [ 1.27095]
MACROECONOMICS AND THE STOCK MARKET 67

STOCK(-3) -0.876597 0.377243 0.225139 3.236176 -0.839818 2.725669 3.329851 2.788109

-0.41284 -7.01476 -0.43408 -1.54031 -9.95129 -3.03802 -5.35394 -1.06258
[-2.12333] [ 0.05378] [ 0.51865] [ 2.10098] [-0.08439] [ 0.89719] [ 0.62194] [ 2.62392]

OIL(-1) 0.007029 0.109144 0.012835 0.015333 -0.404653 0.020475 0.136745 -0.07367

-0.01442 -0.24503 -0.01516 -0.0538 -0.3476 -0.10612 -0.18702 -0.03712
[ 0.48744] [ 0.44544] [ 0.84651] [ 0.28499] [-1.16412] [ 0.19294] [ 0.73119] [-1.98476]

OIL(-2) -0.033417 -0.13021 -0.002658 0.078384 -0.020118 0.034149 -0.0434 -0.01804

-0.01594 -0.27079 -0.01676 -0.05946 -0.38415 -0.11728 -0.20668 -0.04102
[-2.09683] [-0.48085] [-0.15862] [ 1.31824] [-0.05237] [ 0.29119] [-0.20999] [-0.43970]

OIL(-3) 0.019989 0.207833 0.023588 -0.078197 -0.420693 -0.012587 -0.0481 0.019198

-0.01002 -0.17019 -0.01053 -0.03737 -0.24143 -0.07371 -0.12989 -0.02578
[ 1.99573] [ 1.22120] [ 2.23980] [-2.09250] [-1.74249] [-0.17077] [-0.37031] [ 0.74471]

GOLD(-1) 0.892157 50.6793 0.512751 0.139488 -22.39795 -1.876757 2.690872 -0.65184

-0.50317 -8.54963 -0.52906 -1.87734 -12.1287 -3.70275 -6.52541 -1.29507
[ 1.77306] [ 5.92766] [ 0.96917] [ 0.07430] [-1.84669] [-0.50685] [ 0.41237] [-0.50333]

GOLD(-2) 0.075604 5.889043 -0.962735 0.603262 30.10034 -0.226555 2.464167 1.064274

-0.43352 -7.36609 -0.45582 -1.61746 -10.4497 -3.19017 -5.62208 -1.11579
[ 0.17440] [ 0.79948] [-2.11208] [ 0.37297] [ 2.88050] [-0.07102] [ 0.43830] [ 0.95383]

GOLD(-3) -0.996435 -59.54996 -1.442928 -0.175828 35.9675 2.068109 -4.60388 0.473786

-0.39681 -6.74242 -0.41723 -1.48051 -9.56494 -2.92007 -5.14607 -1.02132
[-2.51110] [-8.83214] [-3.45835] [-0.11876] [ 3.76035] [ 0.70824] [-0.89464] [ 0.46390]

HOUSING(-1) 0.255842 4.524995 0.003111 -0.947163 -0.168716 -1.050535 -1.2012 0.003325

-0.15655 -2.65997 -0.1646 -0.58408 -3.77349 -1.152 -2.03019 -0.40292
[ 1.63428] [ 1.70115] [ 0.01890] [-1.62163] [-0.04471] [-0.91192] [-0.59167] [ 0.00825]

HOUSING(-2) -0.080388 -0.674942 -0.044138 0.15752 1.280473 0.225031 0.807521 0.333733

-0.15529 -2.63864 -0.16328 -0.5794 -3.74323 -1.14276 -2.01391 -0.39969
[-0.51766] [-0.25579] [-0.27032] [ 0.27187] [ 0.34208] [ 0.19692] [ 0.40097] [ 0.83497]

HOUSING(-3) -0.154924 1.262603 0.117889 0.74622 -2.339501 0.494961 0.582376 0.096085

-0.07844 -1.33285 -0.08248 -0.29267 -1.89081 -0.57724 -1.01728 -0.2019
[-1.97500] [ 0.94729] [ 1.42933] [ 2.54970] [-1.23730] [ 0.85746] [ 0.57248] [ 0.47591]

GDP(-1) 0.069578 3.119889 0.046029 -0.031569 -2.933573 -0.168372 0.202946 -0.05355

-0.02494 -0.42371 -0.02622 -0.09304 -0.60108 -0.1835 -0.32339 -0.06418
[ 2.79020] [ 7.36334] [ 1.75551] [-0.33931] [-4.88052] [-0.91755] [ 0.62756] [-0.83437]
MACROECONOMICS AND THE STOCK MARKET 68

GDP(-2) 0.07032 4.291286 0.05702 0.037288 -2.604287 -0.22009 0.323848 -0.02341

-0.0356 -0.60485 -0.03743 -0.13281 -0.85805 -0.26195 -0.46164 -0.09162
[ 1.97544] [ 7.09480] [ 1.52341] [ 0.28075] [-3.03512] [-0.84019] [ 0.70151] [-0.25553]

GDP(-3) 0.031631 1.483245 0.004864 -0.011981 -0.611733 -0.115284 0.018635 -0.0202

-0.01721 -0.2925 -0.0181 -0.06423 -0.41495 -0.12668 -0.22325 -0.04431
[ 1.83747] [ 5.07089] [ 0.26871] [-0.18655] [-1.47424] [-0.91005] [ 0.08347] [-0.45585]

INTEREST(-1) -0.058505 -2.465129 0.053445 0.006187 -0.616714 0.713747 0.090313 0.008709

-0.04424 -0.75171 -0.04652 -0.16506 -1.0664 -0.32556 -0.57374 -0.11387
[-1.32242] [-3.27935] [ 1.14893] [ 0.03749] [-0.57832] [ 2.19238] [ 0.15741] [ 0.07648]

INTEREST(-2) 0.092778 2.769792 0.006553 -0.139649 -0.737399 0.040488 -0.07043 -0.05688

-0.0492 -0.83591 -0.05173 -0.18355 -1.18584 -0.36202 -0.638 -0.12662
[ 1.88589] [ 3.31352] [ 0.12668] [-0.76082] [-0.62184] [ 0.11184] [-0.11039] [-0.44924]

INTEREST(-3) -0.027453 0.017117 -0.057883 0.04372 0.983858 -0.191402 0.093953 -0.03656

-0.04351 -0.73931 -0.04575 -0.16234 -1.0488 -0.32019 -0.56427 -0.11199
[-0.63096] [ 0.02315] [-1.26521] [ 0.26932] [ 0.93808] [-0.59778] [ 0.16650] [-0.32648]

EXCHANGE(-1) 0.012978 -0.136081 0.013772 -0.018053 -0.098585 0.331808 -0.4582 0.00407

-0.0267 -0.45369 -0.02807 -0.09962 -0.64361 -0.19649 -0.34627 -0.06872
[ 0.48603] [-0.29994] [ 0.49055] [-0.18121] [-0.15317] [ 1.68870] [-1.32323] [ 0.05923]

EXCHANGE(-2) 0.012469 0.505411 -0.004538 0.028654 -0.077935 0.447918 -0.10619 -0.01568

-0.0306 -0.51987 -0.03217 -0.11415 -0.73749 -0.22515 -0.39678 -0.07875
[ 0.40755] [ 0.97219] [-0.14107] [ 0.25101] [-0.10568] [ 1.98943] [-0.26763] [-0.19906]

EXCHANGE(-3) 0.019341 1.142721 -0.051816 -0.029248 0.48074 -0.216378 -0.09805 -0.02861

-0.0378 -0.64234 -0.03975 -0.14105 -0.91123 -0.27819 -0.49026 -0.0973
[ 0.51161] [ 1.77901] [-1.30360] [-0.20737] [ 0.52757] [-0.77781] [-0.19999] [-0.29400]

CPI(-1) -0.284986 -4.208349 0.044955 0.705861 -1.475084 0.684341 -0.32339 -0.52276

-0.10399 -1.76688 -0.10934 -0.38798 -2.50654 -0.76522 -1.34855 -0.26764
[-2.74060] [-2.38179] [ 0.41115] [ 1.81934] [-0.58849] [ 0.89431] [-0.23980] [-1.95318]

CPI(-2) 0.107983 2.824337 0.218607 0.026973 -6.132227 -0.469845 0.201103 -0.85687

-0.08344 -1.41771 -0.08773 -0.3113 -2.01119 -0.61399 -1.08205 -0.21475
[ 1.29419] [ 1.99219] [ 2.49182] [ 0.08665] [-3.04905] [-0.76523] [ 0.18585] [-3.99009]

CPI(-3) -0.23263 -0.271555 0.129531 0.907585 -3.270209 0.456994 -0.32745 -0.20111

-0.10241 -1.74012 -0.10768 -0.3821 -2.46857 -0.75363 -1.32812 -0.26359
[-2.27152] [-0.15606] [ 1.20292] [ 2.37527] [-1.32474] [ 0.60639] [-0.24655] [-0.76296]

C 0.007642 -0.010177 0.002876 -0.013877 -0.084524 -0.03655 -0.0572 0.002515

-0.00889 -0.1511 -0.00935 -0.03318 -0.21435 -0.06544 -0.11533 -0.02289
[ 0.85933] [-0.06735] [ 0.30756] [-0.41825] [-0.39432] [-0.55853] [-0.49595] [ 0.10987]
MACROECONOMICS AND THE STOCK MARKET 69

The result of the Cholesky normality test presented in Table 19 indicates that the VAR

model data is distributed normally, and that the normality assumption is not violated.

Table 19

VAR Normality Result

Component Skewness χ2 df Prob.

1 -0.05167 0.016909 1 0.8965
2 0.645408 2.638161 1 0.1043
3 0.060731 0.023359 1 0.8785
4 0.227492 0.327767 1 0.567
5 0.294079 0.547724 1 0.4592
6 0.087605 0.048606 1 0.8255
7 -0.571677 2.069825 1 0.1502
8 -0.516206 1.687635 1 0.1939
Joint 7.359985 8 0.4983
Component Kurtosis χ2 df Prob.
1 2.954544 0.003272 1 0.9544
2 3.781187 0.966233 1 0.3256
3 4.038212 1.706651 1 0.1914
4 4.931421 5.906445 1 0.0151
5 4.051537 1.750739 1 0.1858
6 3.043823 0.003041 1 0.956
7 3.662158 0.694218 1 0.4047
8 4.185296 2.224468 1 0.1358
Joint 13.25507 8 0.1034
Note. The null hypothesis of normality cannot be rejected with a probability of
49.83% for skewness and 10.34% for Kurtosis.

In the next section, the IRFs are run, and the results will be discussed separately.

IRF Results

As was discussed earlier, IRF is a very useful tool for VAR which estimates the reaction

of one variable in response to a shock from the other variables within the model. This section

follows the IRF approach for all variables separately. Because the data in this study are

quarterly, this study uses the eight lags to measure the reaction of the variables to external

shocks, which occurred during a period of 2 years. This will facilitate our understanding of the
MACROECONOMICS AND THE STOCK MARKET 70

short-term reaction (within the first two quarters) as well as the long-term reaction (after a lag of

1 or 2 years). In the VAR model, each variable regresses upon the lag of other variables as well

as upon its own lags. This means that most of the time, a portion of the reaction of each variable

is influenced by its own previous lags. This phenomenon is referred to as self-shock. The

reaction of the stock market to its own previous behavior is shown in Figure 11.

Response of STOCK to Generalized One

S.D. STOCK Innovation
.06

.04

.02

.00

-.02

-.04

-.06

-.08
1 2 3 4 5 6 7 8 9 10 11 12

Figure 11. The impulse response of stock market to its own lags.

This figure shows that the stock market strongly responds to its own previous lag almost

immediately after the shock in the first period. It starts approximately 5% higher as soon as the

shock hits the market. However, it moves down sharply below zero to reach a -1.5%

approximately two quarters after the shock. It begins moving slightly higher until approximately

six quarters later, at which time the market reaches positive 1%. After this, the influence of the

shock disappears (after approximately eight quarters).

After this, the study examines the reaction of the stock market to a 1 standard deviation

shock to oil revenue. The results are presented in Figure 12.

MACROECONOMICS AND THE STOCK MARKET 71

Response of STOCK to Generalized One

S.D. OIL Innovation
.08

.06

.04

.02

.00

-.02

-.04

-.06

-.08
1 2 3 4 5 6 7 8 9 10 11 12

Figure 12. The impulse response of the stock market to the oil revenue.

As shown in Figure 12, the reaction of the market to the shock to the oil revenue is not

immediate, but takes approximately one quarter for market to react. The response is slightly

negative and reaches -3% in second quarter. The reaction of the market remains unchanged for

rest of the year (until the fourth quarter). During the second year (fifth quarter) following the

shock (long response analysis) a sharp drop occurred, which was followed by a sharp rise, and

then another drop in the market. This indicates that the market reaction is either influenced by

the response of other macroeconomic reactions to the oil revenue surplus, or depends on how and

where the oil revenue was injected into the economy in general.

After this, the reaction of the stock market price index to a shock to the housing market

activity was examined. The relationship between these two variables was examined in the

previous chapter, and the outcome indicated that housing activity has a significant impact on the

stock market. The result of the stock market reaction to a one standard deviation shock to

housing activity is presented in Figure 13.

MACROECONOMICS AND THE STOCK MARKET 72

Response of STOCK to Generalized One

S.D. HOUSING Innovation
.08

.06

.04

.02

.00

-.02

-.04

-.06

-.08
1 2 3 4 5 6 7 8 9 10 11 12

Figure 13. The impulse response of the stock market to the shock to housing market activity.

As shown in the figure, the reaction of the market to housing activity starts at

approximately - 4% with a sharp positive trend. It rises to approximately .5% during the second

quarter, and then moves slightly move down again, dropping below zero near the end of the year,

followed by a sharp rise starting at the beginning of the second year. The market movement is

very light after the second year, and the indicator stays close to the baseline. Based on the

outcome of the OLS, the next variable with significant impact on the stock market is GDP. The

impulse response to the shock to the GDP is presented in Figure 14.

MACROECONOMICS AND THE STOCK MARKET 73

Response of STOCK to Generalized One

S.D. GDP Innovation
.12

.08

.04

.00

-.04

-.08

-.12
1 2 3 4 5 6 7 8 9 10 11 12

Figure 14. The impulse response of the stock market to the shock to the GDP.

The stock market reacts to GDP during the first quarter of the shock. The market moves

slightly up to approximately 2% at during the second quarter, then drops sharply during the third

quarter. Though it rises slightly, it remains fairly low through the end of the first year. The

indicator continues moving slightly up and down, hovering around .5% during the next four

quarters.

The response of stock market to the other macroeconomic variables is presented in

Figures 15 through 18.

MACROECONOMICS AND THE STOCK MARKET 74

Response of STOCK to Generalized One

S.D. GOLD Innovation
.100

.075

.050

.025

.000

-.025

-.050

-.075

-.100
1 2 3 4 5 6 7 8 9 10 11 12

Figure 15. The impulse response of the stock market to the gold price fluctuation.

Response of STOCK to Generalized One

S.D. INTEREST Innovation
.08

.06

.04

.02

.00

-.02

-.04

-.06

-.08

-.10
1 2 3 4 5 6 7 8 9 10 11 12

Figure 16. The impulse response of the stock market to the long-term interest rate.
MACROECONOMICS AND THE STOCK MARKET 75

Response of STOCK to Generalized One

S.D. EXCHANGE Innovation
.12

.08

.04

.00

-.04

-.08

-.12
1 2 3 4 5 6 7 8 9 10 11 12

Figure 17. The impulse response of the stock market to the foreign currency exchange rate.

Response of STOCK to Generalized One

S.D. CPI Innovation
.06

.04

.02

.00

-.02

-.04

-.06

-.08
1 2 3 4 5 6 7 8 9 10 11 12

Figure 18. The impulse response of the stock market to the CPI.
MACROECONOMICS AND THE STOCK MARKET 76

The reaction of the stock market to the impulse on gold price is the opposite of the

reaction of the market to the impulse on the GDP, while the reaction of the market to the other

macroeconomic variables such as interest rate, foreign exchange, and CPI is relatively small and

remains close to the line zero. However, at the end of the first year after the shock, there is a

small rise followed by a sharp drop (approximately 2%) in the stock market due to a shock to

CPI. The use of the IRFs has helped us arrive at a general understanding of the direction and

timing of the reaction of the stock market to each of the relevant macroeconomic variables

within the model. However, to understand the full significance of the behavior, we must use

another tool in the VAR system known as VDC. The VDC tool provides an estimate what

percentage of the stock market reaction is based on each of the named variables within the

model. The tool is discussed in the next section.

VDC Results

VDC is a tool that gives us information about the percentage of impact of each variable in

the model upon the stock market. In other words, we can estimate what percentage of stock

market behavior is due to impact of oil revenue, what percentage is in response to gold price, and

what percentage is a reaction to other macroeconomic variables. This tool can also measure

these outcomes for different time periods, which makes it possible to understand the both the

short term and long term changes in the impact of variables on each other. The three variables

with the most significant impact on the stock market are oil revenue, housing, and GDP.

Therefore, these three variable were selected for examination in this section of the study, and

were examined at 12 distinct time periods. Table 20 presents the outcome of VDC for these 12

periods for each of the variables.

MACROECONOMICS AND THE STOCK MARKET 77

Table 20

VDC Results

Period S.E. STOCK OIL GOLD HOUSING GDP INTEREST EXCHANGE CPI
1 0.068525 100 0 0 0 0 0 0 0
2 0.080535 76.43908 3.940414 2.636619 0.668097 0.128128 10.8452 2.653758 2.688705
3 0.090529 71.04795 6.545361 6.372693 0.921846 0.112601 10.00796 2.597407 2.394179
4 0.097768 65.74999 6.808021 7.476593 1.724553 1.958005 10.93511 3.279454 2.068273
5 0.117409 54.59208 17.66208 6.255776 1.199314 1.502458 10.07948 5.855366 2.853448
6 0.122903 52.70776 16.15481 9.159698 1.222494 1.543999 10.51017 5.347499 3.353577
7 0.141201 51.06195 16.42977 11.70801 2.014752 1.726156 9.876508 4.491951 2.690902
8 0.145332 48.38129 15.59542 11.12948 2.086519 5.06361 9.660464 5.477558 2.605664
9 0.175929 42.29565 26.13179 7.761631 1.783182 4.661596 8.462315 5.103039 3.80079
10 0.181123 40.21877 27.6651 7.809617 1.704624 5.849933 8.225773 4.918528 3.607654
11 0.214768 36.523 30.39157 8.16656 2.589685 6.696967 7.431165 4.510629 3.690422
12 0.236775 31.09247 34.13733 7.380783 2.772433 10.7025 6.120807 3.720283 4.073395

If we assume the third lag (less than a year) for a measurement of short term and the 12th

lag as a proxy for long term behavior, the table indicates that immediately after the shock, 100%

of the stock market behavior is due to its own previous lags. However, as early as the third lag,

only 71% of stock market behavior can be interpreted as a reaction to the market’s own previous

lags. At this same point in time, approximately 6.5% of its behavior is due to the reaction to the

oil revenue, 0.92% to the housing activity, and 0.11% is due to the impact of the GDP.

However, since we found the coefficient of the GDP to be close to zero, the impact of the GDP

to the stock cannot be relied for the conclusion. Over the long term, at lag 12, only 31% of the

behavior of the stock is due to its own previous lags, 34% is due to oil revenue, 2.78% is due to

the housing market, and 10.7% of its behavior is due to previous lags in GDP. Once more, we

see that the GDP outcome must be disregarded. This is because its impact on the market is close

to zero.
MACROECONOMICS AND THE STOCK MARKET 78

CHAPTER V

Conclusions

This chapter is divided into three distinct sections. The first of these is the summary of

the study, including the research questions and the purpose of the study, a description of the

method used, and descriptions of the models used within the study. The second section reviews

and discusses the findings of the tests conducted in the previous chapter. In the final part,

recommendations for future research will be provided, and the implications of the study will be

discussed.

Summary of the Study

This first section provides the brief summary of the previous chapters, including the

purpose of the study, research questions and hypotheses, the research method, and models which

were used.

Purpose of the Study

During the recent past, there have been some signals from the Iranian economy which

suggest that the Iranian authorities are taking steps toward modernization of the economy and

making efforts to become a part of the global economy; attempting to build a new relationship

with Western countries, specifically with the United States. This may be one of the most

important political moves which can help the Iranian economy move toward modernization.

Such steps may be facilitated by the fact that the younger generation of Iranians is more familiar

with the stock market and with modern theories of finance and economics than are their parents

of any previous generation.

The main purpose of this study has been to examine the behavior of the Iranian stock

market, its relationship with other available investing options, and its role as an investment tool
MACROECONOMICS AND THE STOCK MARKET 79

between Iranian investors. The existing literature has primarily focused on the impact of oil

price on the Iranain stock market (Farzanegan, 2011; Farzanegan & Markwardt, 2007;

Maghyereh, 2004; Mehrara, 2006; Oskooe, 2012) and has generally concluded that the Iranian

stock market does not react significantly to most other macroeconomic variables. Furthermore,

the Iranian stock market has never been studied as an investment tool in comparison to other

common investment markets in Iran, such as gold, real estate, and foreign currency. Due to the

dynamic movement of the Iranian economy through the process of modernization, and due to a

new generation of potential investors in Iran who have much greater knowledge of the finance

and economy than did previous generations, this study has tried to discover the role and the

behavior of the stock market based on these new conditions. This paper also examined the

impact of the selected macroeconomic variables on the volatility of the stock market. In order to

organize the structure of the study, this study focused upon the research questions described in

the following section.

Research Questions

1. Does the oil money inflow4 have any impact on the stock market price index in Iran?

2. Is there any significant relationship between the gold market and the stock market as

two separate investment tools in Iran?

3. Is there any significant relationship between the real estate market and the stock

market index in Iran?

4. Is there any significant relationship between the long term deposit interest rate paid

by banks and the Iranian stock market?

4
Oil money inflow as used in this paper represents Iranian oil revenue instead of oil price fluctuation. This is done
because Iranian oil income might be influenced by political issues such as global sanctions, regardless of the global
price of oil.
MACROECONOMICS AND THE STOCK MARKET 80

5. Is there any significant relationship between foreign currency (U.S. dollar) exchange

rate and the Iranian stock market price index?

6. Would the result of the study have been any different if all of the above variables

were considered within one single model?

Based on the above questions, this study developed the following hypotheses:

1. The oil money inflow fluctuation does significantly impact the Iranian stock market

index price.

2. A higher gold price does impact the stock market index price in Iran.

3. Consideration of oil money inflow and gold price fluctuation in a single model will

change the outcome, and the result will be significantly different.

4. Housing market activity has a direct impact on the stock market index price in Iran.

5. The foreign exchange rate fluctuation as a parallel investment market may have

significant impact on the stock market index price in Iran.

6. GDP and the stock market index price are directly related, and GDP can significantly

impact stock market behavior.

7. CPI has significant impact on the stock market price in Iran.

8. Interest rate offered by banks has a significant impact on the stock market price index

in Iran.

9. The selected macroeconomic variables do have impact on the stock market volatility.

Research Methods and Models

Model 1: The impact of oil revenue on the stock market. Oil income, housing

activity, GDP, long term interest rate, foreign exchange rate, and CPI were selected as

macroeconomic variables to represent the most common investment markets as well as the
MACROECONOMICS AND THE STOCK MARKET 81

general economy of Iran. Selected variables were regressed on the stock market to examine and

analyze any possible relationships. The OLS method was used to develop the model as follows:

𝑆𝑡𝑜𝑐𝑘(𝑡) = 𝛶0 + ∑𝑛𝑗=1 𝛶𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + 𝜉𝑡 (3.4)

where Stock(1t) represents the stock market price index in Iran at time t, ϓ0 is constant, ϓj is the

parameter which needs to be estimated, ξt is the error term, and macroecon is the group of

different macroeconomic variables identified and used to estimate the outcome. The model was

then developed by using the macroeconomic variables, and formulated as follows:

Stockt = ɑ0 + α1oilt-k + α2houset-k + α3exchanget-k + α4gdpt-k + α5cpit-k +

α6Interestt-k + ɛt (3.5)

The main purpose of this model was to examine the relationship of the selected variables

without consideration of the gold market price. In the next model ceteris paribus, the oil revenue

was replaced with the gold market price.

Model 2: The impact of the gold price fluctuation on the stock market. All of the

macroeconomic variables in Model 1 were also used in Model 2, except that oil revenue was

replaced with gold price fluctuation. The purpose of this model was first to examine the

relationship between the selected variables, if any, and secondly, to determine whether the

outcome is different than the outcome of Model 1 due to replacing the oil revenue with the gold

price. The OLS was designed as follows:

𝑆𝑡𝑜𝑐𝑘(𝑡) = 𝜑0 + ∑𝑛𝑗=1 𝜑𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + 𝜔𝑡 (3.6)

where φ0 is constant, φj is the parameter which needs to be determined, ωt is the error term, and

macroecon is a group of macroeconomic variables, such as gold price, GDP, CPI, foreign

exchange rate, and housing activities. By entering the available variables into the model, the

equation can be developed as follows:

MACROECONOMICS AND THE STOCK MARKET 82

Stockt = β0 + β1goldt-k + β2houset-k + β3exchanget-k + β4gdpt-k + β5cpit-k +

β6interestt-k + ωt (3.7)

As mentioned previously, the only difference between Model 1 and Model 2 is that oil

revenue is a variable in Model 1 but not in Model 2, while gold price fluctuation is a variable in

Model 2 but not in Model 1.

Model 3: The impact of oil revenue, the gold price fluctuation, and other

macroeconomic variables on the stock market price index. This model was developed to

cover all selected macroeconomic variables, including oil revenue and gold price fluctuation.

There were two purposes for developing this model. The main purpose of this model was to

discover the relationship between the selected macroeconomic variables and the stock market

price index. The second purpose was to investigate whether consideration of gold price as a

proxy for the gold investment market along with oil revenue (and other macroeconomic

variables) would change the outcome of the model. Because most of the previous literature has

focused on the oil price and stock market, the intent behind the creation of this model was to test

investment in gold as a parallel investment market to the stock market. The model was designed

as follows:
𝑛
𝑆𝑡𝑜𝑐𝑘(𝑡) = Ӯ0 + ∑𝑗=1 Ӯ𝑗 𝑚𝑎𝑐𝑟𝑜𝑒𝑐𝑜𝑛 + Ӫ𝑡 (3.8)

where Stock(t) is the stock market activity in Iran, Ӯ0 is constant, Ӯj is the variable to be

determined, Ӫt is the error term, and macroecon is the group of macroeconomic variables which

are identified in the data description. After inserting the variables into the model, the final model

was developed as follows:

Stockt= χ0 +χ1oilt-k +χ2goldt-k +χ3houset-k +χ4exchanget-k +χ5gdpt-k +χ6cpit-k +

χ7interestt-k + ςt (3.9)
MACROECONOMICS AND THE STOCK MARKET 83

where χ0 is constant, χ1 through χ7 are parameters to be estimated, ςt is the random error term, k

is the proper lag, and stock, oil, gold, house, exchange, dgp, cpi, and interest represent the first

difference of the logarithm of the Iranian stock market activity, which are oil revenue, gold price,

housing activities, foreign exchange rate, GDPs, CPI and long-term deposit interest rate,

respectively.

Model 4: The impact of oil revenue, the gold price fluctuation and other

macroeconomic variables on stock market volatility. After examining the relationship

between all macroeconomic variables and the stock market price index, this model was

developed to discover any possible impact of the variables on the volatility of the stock market.

Volatility was measured as the fluctuation of the annual return of the stock market. The model

was designed as follows:

Volatilityt= Ө0 +Ө1oilt-k +Ө2goldt-k +Ө3houset-k +Ө4exchanget-k +Ө5gdpt-k

+Ө6cpit-k + Ө7interestt-k + Ωt (3.10)

where Ө0 is constant, Ө1 through Ө7 are coefficients to be determined, oil, gold, house, gdp, cpi,

and interest are Iranian oil revenue, gold price, housing activities, foreign exchange rate, GDPs,

CPI, and long-term deposit interest rate respectively, and Ωt is the error term.

Discussion and Analysis of the Research Findings

This section discusses the result of each model and the findings about each individual

hypothesis.

Hypothesis 1: The oil money inflow fluctuation will significantly impact the Iran

stock market index price. Current literature in general reports that there is no correlation

between oil price and the Iranian stock market price index. This paper used the oil revenue to

measure the possible impact because, regardless of global oil price fluctuations, Iranian oil
MACROECONOMICS AND THE STOCK MARKET 84

revenue is influenced by external factors such as global sanctions, global oil demand, and

Organization of the Petroleum Exporting Countries regulations. The result of the first model

indicates that among all variables, oil revenue, housing activity, and GDP each significantly

impacts the stock market price index. However, the oil coefficient was only 1.4%, thus

demonstrating its impact is very weak. In addition, the GDP coefficient is essentially zero

(-0.1%), and therefore its impact is sufficiently minimal that it can be disregarded. Housing

market activity, with a coefficient of -22.6%, has the greatest impact on the stock market price.

That means when the housing activity is down, the stock price index rises higher. The R2 of the

model is 87.5%, which indicates that the model can explain more than 87% of the variation

within the stock market. This finding may suggest that housing activity may be a substitute

investment for the stock market in Iran. The result of Model 3, which included all variables (as

well as gold price) was the same for the three significant variables.

Hypothesis 2: The higher gold price will impact the stock market index price in

Iran. Gold was examined first in Model 2, along with other variables, but not with oil revenue.

After this, gold price was examined in Model 3 along with all other variables. The null

hypothesis, which says that gold price does not have any impact on the stock price index in Iran,

cannot be rejected, neither in Model 2 with a probability of 46%, nor in Model 3 with a

probability of 59.81%. This finding indicates that not only is gold not a substitute of investment

for stock market, but no other correlation could be find between the gold prices and the stock

market. There may be several reasons for this finding. For example, the investors in gold and in

the stock market may be from two different levels of the society. In other words, people from a

certain socioeconomic segment of the society may be interested in investing in stock market,

while other people in different classes or segment of the society may prefer to invest in the gold
MACROECONOMICS AND THE STOCK MARKET 85

market. Another possibility is that gold price may not be the best proxy to measure the gold

market activity in Iran. Unfortunately, at the time of preparation of this study, gold price was the

only available proxy for the gold market in Iran.

Hypothesis 3: Consideration of oil money inflow and gold price fluctuation in a

single model will change the outcome, and the results may be significantly different. The

outcome of the three models is essentially the same. That means that, regardless of consideration

the gold price in the model, the impact of the variables on the stock market was more or less the

same. Furthermore, gold did not have any impact on the stock market at all. In all of the

models, housing activity had a strong and negative impact on the stock market, but oil revenue

(positive 1.4%) and GDP (negative 0.1%) each had only a very weak coefficient, and therefore,

the impact of both variables can be disregarded. The only noteworthy difference between Model

2 and Model 3 is the housing activity coefficient, which was greater in Model 3 (-22.52%) than

in Model 2 (-18.64%).

Hypothesis 4: Housing market activity has a direct impact on the stock market index

price in Iran. With a zero probability in all models, the null hypothesis, which claims that

housing activity has no impact on the stock market price index in Iran, can be rejected.

Conversely, the alternative hypothesis which claims the housing activity has a direct impact on

the stock market can be accepted. The housing market activity had the highest impact (greater

than -18% in Model 2, and greater than -22% in Model 1 and Model 3) on the stock market price

index. That finding indicates that housing market activity in Iran can be considered as a

substitute investment for the stock market. An explanation for this outcome may be that when

they think that the stock return is lower or riskier for any reason, these private investors who
MACROECONOMICS AND THE STOCK MARKET 86

normally invest in the stock market pull their capital out of the stock market and invest it in real

estate instead.

Hypothesis 5: The foreign exchange rate fluctuation as a parallel investment market

has a significant impact on the stock market index price in Iran. The foreign exchange rate

was not significant in any of the models in this study. With a probability of greater than 73%,

the null hypothesis, which states the foreign exchange rate has no impact on stock market

activity, cannot be rejected. In other words, based on the findings in this paper, there is no

correlation between the stock market price index and the foreign exchange rate, so therefore

foreign currency cannot be considered as an alternative investment or a substitute for the stock

market.

Hypothesis 6: GDP and the stock market index price are directly related, and GDP

can significantly impact the stock market behavior. With a probability of less than 0.1%, and

approximately zero in all three models, the null hypothesis stating that the GDP has no impact on

the stock market price index can be rejected. GDP had a negative coefficient in the model.

However, in all cases, the coefficient of GDP is very small (0.1% in Model 1 and 2, and nearly

zero in Model 3), and the impact is very weak. Based on the outcome numbers, the impact is

effectively zero, and therefore can be disregarded. That means that, despite being statistically

significant, GDP has no substantial impact on the stock market price index in Iran.

Hypothesis 7: CPI has a significant impact on the stock market price in Iran. CPI

had one of the highest statistical probabilities in Models 1, 2, and 3. That means that the CPI

null hypothesis cannot be rejected, and CPI does not have any impact on the stock market price

index in Iran.
MACROECONOMICS AND THE STOCK MARKET 87

Hypothesis 8: Interest rate offered by banks has a significant impact on the stock

market activities in Iran. As with the foreign currency exchange rate, the interest rate also had

a very high statistical probability, and therefore the null hypothesis cannot be rejected. The 5-

year deposit interest rate offered by banks is relatively high in Iran (sometimes more than 20%).

The fact that this study could not find any correlation between interest rate and stock market

price index can be interpreted based on the two facts: either because the return on the stock

market is lower than the interest rate, or because that it is riskier than depositing money in banks,

or a combination of both.

Hypothesis 9: The selected macroeconomic variables have impact on the stock

market volatility in Iran. The outcome of Model 4, which measures the relationship between

each of the variables and the stock market volatility, indicates that oil revenue, gold price,

housing activity, and CPI each do have an impact on the volatility of the stock market, but that

GDP, interest rate, and foreign exchange rate do not. Among all these variables, only housing

has positive impact on the stock market volatility, but oil revenue, gold price, and CPI have a

negative impact. Housing activity also has the strongest coefficient with regards to the stock

market volatility (16%). This indicates there is a direct and strong correlation between the stock

market volatility and housing market activity.

The finding in this study was in contrast with some of previous papers regarding Iran

stock market and other macroeconomics in which they found oil, gold, and GDP to be all

significantly impacting the stock market. The reason may be based on the fact that first of all the

other paper used the row data instead of the log level and secondly because they manipulated the

regression process by adding one lag of the stock market to the regression process. This can

change the whole result and gives us the spurious result since the regression method already
MACROECONOMICS AND THE STOCK MARKET 88

analyzes the data based on their own previous lags. Considering the fact that large percentage of

the stock market behavior is based on its own previous lag, adding one lag to the regression

analysis would create the spurious outcome.

In this study, macroeconomic variables were selected as proxies of different common

investment methods, as well as indicators for the whole economy in Iran. The Iranian economy

is taking steps towards modernization, and the Iranian population as a whole is becoming more

familiar with the stock market as a modern tool of investment. However, the fact that only one

indicator (housing activity) has significant relationship with the stock market indicates that the

stock market cannot yet be considered as a common investment tool in Iran.

This seems like an interesting finding. As mentioned before, there may be several

explanations for that phenomenon. First, lack of knowledge and familiarity among domestic

investors about the stock market; second, lack of trust to the stock market investment; third,

lower rate of return of the stock market comparing to the other investments tools; and fourth,

higher risk on the stock market investment. The problem of lack of knowledge about the stock

market seems to be improving since the new generation of Iran population is more educated

about the market. However, trust to the stock market system among investors as well as rate of

return of the market still seems to need more improvements. Because the Iran stock market is a

relatively new system of investment, the process of providing information about the market still

needs improvement. For example, there are only very limited year-end annual reports available

on the TSE website of which the volume and the nature of the provided information as well as

the format of the reports are different each year. Out of those limited number of reports, the

stock market authorities provided the rate of return of the market only for 3 years as follow:

13.92% for 2007, 53.71% for 2010, and 29.5% for 2011. They addressed the huge rate of return
MACROECONOMICS AND THE STOCK MARKET 89

of the 2010 to the increase of the share price of the construction sector. However, two things

need to be considered regarding these numbers. First, lack of transparency can make the distrust

of investors to the market even worse. Because only 3 years of data regarding rate of return of

the market is available on the website, it can send a signal to investors that the system is not

reliable. Also, the difference between numbers for different years is huge which can add to the

confusion. The fact that there’s not exist an organization like Security of Exchange Committee

to supervise the market and to guarantee that the provided information is reliable can make the

problem even worse. The other thing to be considered is the fact that the provided rate of return

is not the real rate of return. It should be adjusted for inflation and inflation is a huge problem in

Iran. For example, Iran reported an inflation during 2006 through 2008 as of 11.87%, 18.38%,

and 25.38%, respectively. During the period of study, Iran also experienced the highest inflation

on 1995 of 49.11%. After adjusting for these numbers, the high risk of inflation (which is almost

unpredictable in Iran) and the lower real rate of return of the market prevent investors of

deciding to invest in the Iran stock market.

By analyzing the above data, it can be easier to understand why investors prefer to invest

in real estate rather than stock market in Iran. Almost all real estate transaction in Iran occurs as

cash purchase. It is not common that banks offer loan packages for homebuyers in Iran. Even in

very limited cases if any bank offers loan for that purpose, the amount of the loan is very small

percentage of the whole price of the home and is not considerable. As a result, the price of real

estate in Iran is not influenced by interest rate since the owner does not owe any principal or

interest to banks. That’s why real estate price in Iran almost always has an upward trend. Even

when the market is down for any reason, the only impact is that number of transactions would be

reduced for a period of time; but the price almost never fall. So, from investing point of view,
MACROECONOMICS AND THE STOCK MARKET 90

investment on real estate is guaranteed and the possibility of depreciation of the value of

investment is minimal.

The question of why there is not a strong relationship between other selected

macroeconomic variables and the stock market is also not easy to answer. This question needs to

be the focus of future research. However, at some theories might be generated based on the

finding of this study. For example, the payment of high interest rates by banks might be one

reason why people prefer to deposit their money in banks instead of investing in the stock

market. Because there is a high inflation rate in Iran, the expectation of the rate of return is also

high. Bank deposits are also subject to the risk of inflation, but, by offering the high rate bay

banks, that risk is also eliminated. In other words, banks are competing with the stock market to

attract investors’ money. Banks are apparently doing so with a high degree of success, for it is

evidently difficult for stocks to compete with the rates of return that are offered by banks.

In order to create a better-functioning stock market in the future, Iran has a great need to

attract foreign investors. Regardless of the return on the market (which is, of course, very

important for investors), the security of investments in Iran will need to be guaranteed in the

future better than they are today. These important steps seem to be feasible only by forming a

much closer economic and political relationship with the Western world in the future than has

existed in the recent past or that exists at the present time.

Future Research Recommendations

As mentioned previously in this paper, the stock market is a relatively new investment

tool in Iran. The Iranian stock market still is not very well established or well accepted among

domestic investors. In addition, the economy of Iran seems to be in the process of taking steps

toward modernization, and seems to be working towards building a closer relationship with
MACROECONOMICS AND THE STOCK MARKET 91

Western countries. As a result, the relationship between the stock market and behavior of

investors is dynamic, and might well change considerably with the passage of time. Any future

research in this field might obtain quite different results than reported by this study, depending

future actions of the Iranian government or changes within Iranian society.

I would suggest the following questions should be studied in the future to help achieve a

better understanding of the behavior of the Iranian stock market.

1. Among all macroeconomic variables, why does the real estate market have such a

strong link with the stock market price index?

2. What is the relationship between the economy of Iran in general and foreign

investment, if any?

3. What is the role of the Central Bank of Iran in preparing a better economic

environment in which the stock market can expand and flourish?

4. The Iranian economy is severely dependent upon Iranian oil income. What steps

need to be taken by the government and by private sectors to shift the economy from

being completely oil revenue-based towards a more diversified economy in which the

stock market plays an important role in providing the capital for businesses?

5. How can mutual funds like REITs and Gold Traders be established in the Iranian

stock market, and thereby make it possible for individuals and companies to invest in

gold and real estate through the Iranian stock market?

MACROECONOMICS AND THE STOCK MARKET 92

References

Amuzegar, J. (1997). Iran’s economy under the Islamic Republic. London, United Kingdom:

Tauris.

Arouri, M. & Lahiani, A. & Bellalah, M. (2010). Oil price shocks and stock market returns in

oil-exporting countries: The case of GCC countries. International Journal of Economics

and Finance, 2, 132–139. doi:10.5539/ijef.v2n5p132

Asteriou, D., & Bashmakova, Y. (2009) Assessing the impact of oil returns on emerging stock

markets: A panel data approach for ten Central and Eastern European Countries. Energy

Economics, 38, 204–211. doi:10.1016/j.eneco.2013.02.011

Basher, S. (2014, January 5). Stock market and energy prices. Retrieved from

http://www.syedbasher.org/published/MS05548.pdf

Baur, D., & Lucey, B. (2010). Is gold a hedge or a safe haven? An analysis of stock, bonds and

gold. The Financial Review, 45, 217–229. Retrieved from

https://financialreview.poole.ncsu.edu/

Bjornland, H. (2008). Oil price shocks and stock market booms in an oil exporting country.

Retrieved from http://home.bi.no/a0310125/SJPE_OilPrice.pdf

Central Bank of Iran. (n.d.). Time series database. Retrieved from http://tsd.cbi.ir/

Chen, S. (2010). Do higher oil prices push the stock market into bear territory? Energy

Economics, 32, 490–495. doi:10.1016/j.eneco.2009.08.018

Cologni, A. & Manera, M. (2005). Oil prices, inflation and interest rates in a structural

cointegrated VAR Model for the G-7 Countries. International Energy Markets, 30, 856–

888. doi:10.1016/j.eneco.2006.11.001
MACROECONOMICS AND THE STOCK MARKET 93

Davis, S., & Haltiwanger, J. (2001). Sectoral job creation and destruction responses to oil price

changes. Journal of Monetary Economics, 48, 465–512. doi:10.1016/S0304-

3932(01)00086-1

Dhaoui, A., & Khraief, N. (2014). Empirical linkage between oil price and stock market returns

and volatility: Evidence from international developed markets (Discussion Paper No.

2014-12). Retrieved from http://www.economics-ejournal.org/economics

/discussionpapers/2014-12

Farzanegan, M. (2011). Oil revenue shocks and government spending behavior in Iran. Energy

Economics, 33, 1055–1069. doi:10.1016/j.eneco.2011.05.005

Farzanegan, M., & Markwardt, G. (2007). The effects of oil price shocks on the Iranian

economy. Energy Economic, 31, 134–151. doi:10.1016/j.eneco.2008.09.003

Guiso, L., & Sapienza, P., & Zingales, L. (2008). Trusting the stock market. Journal of Finance,

63, 2557–2600. doi:10.1111/j.1540-6261.2008.01408.x

Gilmore, C., & McManus, G., & Sharma, R., & Tezel, A. (2009). The dynamic of gold price,

gold mining stock price, and stock market prices co-movements. Research in Applied

Economics, 1(1), E12. Retrieved from http://www.macrothink.org/journal/index.php/rae

Hamilton, J. (1983). Oil and the macro-economy since WWII. Journal of Political Economy, 91,

228–248. doi:10.1086/261140

Jiménez-Rodríguez, R., & Sanchez, M. (2004). Oil price shocks and real GDP growth:

Empirical evidence from some OECD countries (Working Paper No. 362). Retrieved

from European Central Bank website: https://www.ecb.europa.eu/pub/pdf/scpwps

/ecbwp362.pdf?b35d2a5fd0bae52378b274ce13a956c4
MACROECONOMICS AND THE STOCK MARKET 94

Kilian, L. (2009). Not all oil price shocks are alike: Disentangling demand and supply shocks in

the crude oil market. American Economic Review, 99, 1053–1069.

doi:10.1257/aer.99.3.1053

Le, T. H., & Chang, Y. (2011, June 16). The impact of oil price fluctuations on stock markets in

developed and emerging economies. Retrieved from https://mpra.ub.uni-

muenchen.de/31753/1/MPRA_paper_31753.pdf

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. Retrieved from http://www.usc.es/economet/ijaeqs.htm

Mehrara, M. (2006). The relationship between stock market and macroeconomic variables: A

case study of Iran. Iranian Economic Review, 11, 138–148. Retrieved from

https://ier.ut.ac.ir/

Mehrara, M., & Sarem, M. (2009). Effects of oil price shocks on industrial production: Evidence

from some oil-exporting countries. OPEC Energy Review, 33, 170–183.

doi:10.1111/j.1753-0237.2009.00167.x

Mishra, P., & Das, K., & Mishra, S. (2010). Gold price volatility and stock market returns in

India. American Journal of Scientific Research, 9, 47–55. Retrieved from

http://www.americanjournalofscientificresearch.com/

Monjazeb, M., & Hoseinzadeh, H. (2013). Impact of oil exports income on Tehran Stock

Exchange Dividend and Price Index (TEDPIX), model selection. World Applied Sciences

Journal, 22, 819–823. doi:10.5829/idosi.wasj.2013.22.06.210

MACROECONOMICS AND THE STOCK MARKET 95

Oskooe, S. A. P. (2012). Oil price shocks and stock market in oil-exporting countries: Evidence

from Iran stock market. OPEC Energy Review, 36, 396–412. doi:10.1111/j.1753-

0237.2012.00217.x

Park, J., & Ratti, R. (2007). Oil price shocks and stock markets in the U.S. and 13 European

Countries. Energy Economics, 30, 2587–2608. doi:10.1016/j.eneco.2008.04.003

Rad, A. A. (2011). Macroeconomic variables and the stock market: Evidence from Iran.

International Journal of Economic and Finance Studies, 3(1), 1–10. Retrieved from

http://www.sobiad.org/ejournals/journal_IJEF/index.html

Raraga, F., & Muharam, H. (2013). VAR analysis on mutual relationship between stock price

index and exchange rate and the role of world oil price and world gold price. SSRN

Electronic Journal. doi:10.2139/ssrn.2470184

Sahu, N., Bandopadhyay, K., & Mondal, D. (2014). An empirical study on the dynamic

relationship between oil prices and Indian stock market. Managerial Finance, 40, 200–

215. doi:10.1108/MF-06-2013-0131

Sadorsky, P. (1999). Oil price shocks and stock market activity. Energy Economics, 21, 449–

469. doi:10.1016/S0140-9883(99)00020-1

Shafiee, S., & Topal, E. (2010). An overview of global gold market and gold price forecasting.

Resource Policy, 35, 178–189. doi:10.1016/j.resourpol.2010.05.004

Sims, C. (1980). Macroeconomics and reality. Econometrica, 48, 1–48. Retrieved from

http://www.econometricsociety.org/publications/econometrica/browse

Smith, G. (2001, November). The price of gold and stock price indices for the United States.

Retrieved from http://www.spdrgoldshares.com/media/GLD/file

/Gold%26USStockIndicesDEC200120fina.pdf
MACROECONOMICS AND THE STOCK MARKET 96

Tehran Stock Exchange. (n.d.). History. Retrieved from http://new.tse.ir/cms/page56.html

Wang, Y., & Wu, C. & Yang, L. (2012). Oil price shocks and stock market returns: Evidence

from oil-importing and oil-exporting countries. Journal of Comparative Economics, 41,

1220–1239. doi:10.1016/j.jce.2012.12.004

Yahyazadehfar, M., & Babai, (2012). Macroeconomic variables and stock price: New evidence

from Iran. Middle Journal of Scientific Research, 11, 408–415. Retrieved from

http://www.idosi.org/mejsr/online.htm
MACROECONOMICS AND THE STOCK MARKET 97

APPENDIX

Title of Appendix
MACROECONOMICS AND THE STOCK MARKET 98

Table A1

Log Level of the Raw Data of All Variables

Date STOCK OIL GOLD HOUSING INTEREST EXCHANGE GDP CPI

1990Q1 NA NA NA NA NA NA NA NA
1990Q2 NA NA NA NA 0.25 NA NA NA
1990Q3 NA 0.00888875 NA -0.01361625 0.25 -0.0516 NA 0.01
1990Q4 NA 0.00888875 NA -0.01361625 0.25 0.2233 NA 0.03
1991Q1 NA 0.00888875 -0.0022 -0.01361625 0.25 0.0125 NA 1.11E-16
1991Q2 NA 0.00888875 0.0017 -0.01361625 0.25 -0.0836 -0.114333333 0
1991Q3 3.37E-05 0.00888875 -0.0131 -0.01361625 0.25 -0.0244 0.283333333 0
1991Q4 3.38E-05 0.00888875 0.0064 -0.01361625 0.25 -0.0357 -0.438333333 0.03
1992Q1 3.37E-05 -0.092964375 0.0067 -0.01329125 0.25 -0.0303 0.173 0
1992Q2 3.38E-05 0.151483125 -0.0053 -0.01407125 0.25 0.0577 0.123666667 -0.05
1992Q3 3.37E-05 0.04963 -0.0025 -0.01374625 0.25 0.0383 0.017666667 0.01
1992Q4 3.38E-05 0.04963 0.0033 -0.01374625 0.25 -0.0705 -0.29 0.04
1993Q1 -0.004920938 0.21653 0.0018 -0.041930625 0.40625 0.035 0.359333333 -2.22E-16
1993Q2 0.006970313 -0.18403 0.0038 0.025711875 0.1875 -0.0241 -0.344333333 -0.04
1993Q3 0.002015625 -0.01713 0.0069 -0.0024725 0.125 0.0621 0.155666667 0.04
1993Q4 0.002015625 -0.01713 -0.007 -0.0024725 0.0625 0.0187 0.425333333 0.05
1994Q1 -0.001759375 -0.078001875 0.0029 -0.010158437 -0.546875 -0.0083 -0.506333333 -2.22E-16
1994Q2 0.007300625 0.068090625 0.0535 0.008287812 0.15625 -0.1142 -0.969333333 -0.04
1994Q3 0.003525625 0.00721875 -0.0166 0.000601875 0.3125 0.2054 2.058 0.06
1994Q4 0.003525625 0.00721875 -0.0386 0.000601875 0.46875 0.307 -0.690333333 0.09
1995Q1 0.019645938 0.013229688 -0.0033 0.02353625 1.40625 -0.0869 0.55 0.15
1995Q2 -0.019042813 -0.001196563 0.1431 -0.03150625 0.46875 -0.5252 -4.997 -0.31
1995Q3 -0.0029225 0.004814375 -0.0452 -0.008571875 0.3125 0.3789 7.242333333 0.1
1995Q4 -0.0029225 0.004814375 -0.1922 -0.008571875 0.15625 -0.1007 -1.650333333 0.04
1996Q1 0.000257188 0.179789375 0.0678 -0.020810938 -0.390625 1.578 -3.235666667 -8.88E-16
1996Q2 -0.007374063 -0.240150625 0.0384 0.008562813 NA -2.4971 0.854666667 -0.19
1996Q3 -0.004194375 -0.065175625 -0.0155 -0.00367625 NA 0.9239 2.702333333 0.13
1996Q4 -0.004194375 -0.065175625 -0.0179 -0.00367625 NA 0.0521 -2.440666667 -0.03
1997Q1 -0.017314687 -0.037327187 0.0298 -0.057241875 -3.55E-15 -0.053 0.101333333 0.1
1997Q2 0.014174063 -0.104163438 -0.01 0.071315625 NA -0.0815 1.262667 -0.16
1997Q3 0.00105375 -0.076315 -0.0087 0.01775 NA 0.3249 -0.829333333 0.02
1997Q4 0.00105375 -0.076315 -0.0108 0.01775 3.55E-15 -0.3132 0.121333333 0.15
1998Q1 -0.008550938 -0.822682187 0.0051 -0.035842188 -3.55E-15 -0.0028 0.311 -8.88E-16
1998Q2 0.014500313 0.968599062 0.013 0.092779062 -3.55E-15 -0.0899 -1.013666667 -0.15
1998Q3 0.004895625 0.222231875 0.0259 0.039186875 3.55E-15 0.0156 1.265 0.08
1998Q4 0.004895625 0.222231875 -0.0203 0.039186875 3.55E-15 0.2239 -0.411333333 0.09
1999Q1 0.01551125 0.8772225 0.0507 0.1862775 -3.55E-15 0.4223 -0.680666667 0.15
1999Q2 -0.00996625 -0.694755 0.0021 -0.16674 NA -0.3717 0.097333333 -0.37
1999Q3 0.000649375 -0.039764375 -0.0427 -0.019649375 NA 0.6247 2.053666667 0.13
1999Q4 0.000649375 -0.039764375 0.0099 -0.019649375 3.55E-15 0.1713 -2.763666667 0.12
2000Q1 0.002141563 -0.095901875 0.0188 -0.041372813 0.234375 -0.5164 1.27 -0.11
2000Q2 -0.001439688 0.038828125 -0.0259 0.010763437 -0.09375 0.191 -0.105666667 -0.1
2000Q3 5.25E-05 -0.017309375 -0.0216 -0.01096 -0.1875 -1.092 0.409333333 0.09
2000Q4 5.25E-05 -0.017309375 -0.0189 -0.01096 -0.28125 0.0893 -0.633 0.05
MACROECONOMICS AND THE STOCK MARKET 99

2001Q1 -0.007995937 -0.34434375 0.0047 -0.107361562 -0.84375 0.2678 0.557333333 -0.01

2001Q2 0.011320312 0.44053875 0.0311 0.124002187 -0.28125 -0.1861 -1.398 -0.1
2001Q3 0.003271875 0.113504375 -0.0095 0.027600625 -0.1875 0.059 2.116666667 0.03
2001Q4 0.003271875 0.113504375 -0.0216 0.027600625 -0.09375 0.0287 -1.368 0.15
2002Q1 -0.066870312 0.477605937 0.0382 0.18736 0.234375 0.1117 0.208333333 0.09
2002Q2 0.101470937 -0.396237812 0.0375 -0.1960625 NA -0.0243 -0.163 -0.17
2002Q3 0.03132875 -0.03213625 0.0031 -0.036303125 NA 0.0298 1.006333333 -0.01
2002Q4 0.03132875 -0.03213625 0.0001 -0.036303125 NA -0.0316 -1.541666667 0.23
2003Q1 0.196885 -0.061412813 0.0022 -0.005567187 NA 0.0057 NA -0.03
2003Q2 -0.20045 0.008850937 0.0724 -0.079333438 NA 0.0183 -4.063666667 -0.21
2003Q3 -0.03489375 -0.020425625 -0.0659 -0.0485975 NA 0.0066 5.888333333 0.01
2003Q4 -0.03489375 -0.020425625 -0.0135 -0.0485975 NA 0.0216 -4.459666667 0.18
2004Q1 -0.069203125 -0.28507875 0.0164 -0.38451625 NA 0.0902 -4.41201 0.11
2004Q2 0.013139375 0.35008875 0.0307 0.42168875 NA 0.0118 -0.974333333 -0.19
2004Q3 -0.02117 0.085435625 -0.0375 0.08577 NA -0.0832 2.134333333 -0.05
2004Q4 -0.02117 0.085435625 -0.0066 0.08577 NA -0.0457 -1.389 0.13
2005Q1 -0.1212075 0.930123125 0.0752 0.46822625 0.15625 0.1821 -1.727 0.09
2005Q2 0.1188825 -1.097126875 0.027 -0.44966875 -0.0625 -0.0436 1.699333333 -0.57
2005Q3 0.018845 -0.252439375 -0.1286 -0.0672125 -0.125 -0.0846 2.902333333 0.32
2005Q4 0.018845 -0.252439375 -0.0065 -0.0672125 -0.1875 0 -3.924666667 0.12
2006Q1 0.067530937 -0.82258625 0.0281 -0.350953125 -0.5625 0.0096 1.399666667 0.03
2006Q2 -0.049315312 0.54576625 0.1957 0.330024375 -0.1875 0.0093 -3.692666667 -0.08
2006Q3 -0.000629375 -0.024380625 0.0716 0.04628375 -0.125 -0.0282 4.084666667 0.23
2006Q4 -0.000629375 -0.024380625 -0.2944 0.04628375 -0.0625 -0.002 3.972333333 0.13
2007Q1 0.034926875 -0.87429 0.0065 0.057693125 -0.3125 -0.0159 -7.137333333 -0.13
2007Q2 -0.050408125 NA 0.0874 0.030310625 0.1875 0.0017 -1.117 -0.08
2007Q3 -0.014851875 0.315583125 -0.0831 0.04172 0.375 -0.0029 9.501666667 0.36
2007Q4 -0.014851875 0.315583125 -0.049 0.04172 0.5625 0.0057 -6.826666667 0.19
2008Q1 -0.156444062 NA 0.2783 0.22081375 1.921875 -0.0217 -1.778666667 0.27
2008Q2 0.183377187 -2.975670313 0.1822 -0.20901125 0.46875 0.0203 -4.734666667 -0.36
2008Q3 0.041785 -0.624775 -0.2928 -0.0299175 0.1875 0.0086 21.83633333 0.16
2008Q4 0.041785 -0.624775 -0.2484 -0.0299175 -0.09375 -0.0651 -20.623 -0.94
2009Q1 0.04956 NA 0.1387 0.168034063 -1.234375 -0.1042 5.445666667 0.41
2009Q2 0.0309 8.8791425 -0.0993 -0.307049688 -0.3125 0.3492 2.244666667 -0.03
2009Q3 0.038675 2.09063 0.057 -0.109098125 -0.25 0.3811 -1.476666667 -0.19
2009Q4 0.038675 2.09063 0.1254 -0.109098125 -0.1875 -0.8419 -0.252333333 0.13
2010Q1 0.262667187 9.54964875 0.2449 -0.579521562 -0.515625 0.2717 -4.558 0.18
2010Q2 -0.274914062 -8.351996 -0.1219 0.549494687 0.09375 0.0472 11.28133 0
2010Q3 -0.050921875 -0.8929775 -0.0794 0.07907125 0.3125 -0.119 -8.119333333 0.28
2010Q4 -0.050921875 -0.8929775 0.1136 0.07907125 0.53125 0.1322 1.542666667 0.74
2011Q1 -0.327046875 NA 0.2343 0.003179062 1.765625 0.1046 -1.208 -0.23
2011Q2 0.335653125 -3.816916875 0.0108 0.185320313 0.5625 0.0496 -2.109333333 -0.61
2011Q3 0.059528125 -1.72838875 -0.0762 0.109428125 0.375 -0.057 17.71933333 0.46
2011Q4 0.059528125 -1.72838875 0.1006 0.109428125 0.1875 -0.0039 -33.816 0.6
2012Q1 0.059528125 -1.72838875 0.4956 0.109428125 0 0.4235 40.759 NA
2012Q2 0.059528125 -1.72838875 0.5924 0.109428125 -0.1875 -0.3641 -71.50466667 NA
2012Q3 0.059528125 -1.72838875 -2.2661 0.109428125 -0.375 1.0974 153.5333333 NA
2012Q4 0.059528125 -1.72838875 1.4789 0.109428125 -0.5625 3.4425 -204.2133333 NA
MACROECONOMICS AND THE STOCK MARKET 100

Table A2

Descriptive Statistics of Variables

STOCK OIL GOLD HOUSING INTEREST EXCHANGE GDP CPI

Mean 0.006814 0.018109 -0.003022 0.005487 0.053421 0.000558 -0.294842 0.021081

Median 0.000649 -0.01722 0.00195 -0.006124 0 0.0076 0.0575 0.035
Maximum 0.335653 9.549649 0.2449 0.549495 1.40625 1.578 21.83633 0.74
Minimum -0.274914 -8.351996 -0.2944 -0.579522 -0.84375 -2.4971 -33.816 -0.94
Std. Dev. 0.073823 1.987466 0.088897 0.161607 0.313504 0.437431 6.501073 0.237692
Skewness 0.869649 1.574285 -0.7296 -0.003525 0.694956 -2.065906 -1.381248 -0.794321
Kurtosis 11.98197 17.52099 6.124894 7.487995 6.962784 18.28857 14.16026 7.192281

Jarque-Bera 258.0777 680.7161 36.67387 62.10497 54.37617 773.338 407.5636 61.97193

Probability 0 0 0 0 0 0 0 0

Sum 0.504259 1.340049 -0.2236 0.406038 3.953125 0.0413 -21.81834 1.56

Sum Sq. Dev. 0.39784 288.3515 0.576892 1.906527 7.174808 13.96823 3085.268 4.124314

Observations 74 74 74 74 74 74 74 74