Thesis 8-2-16.docx

Embed Size (px)

Citation preview

  • 7/26/2019 Thesis 8-2-16.docx

    1/129

    OIL PRICE DETERMINANTS AND CO-MOVEMENT DYNAMICS

    A THESIS

    submitted by

    SRINIVASAN.N

    for the award of the degree

    of

    MASTER OF SCIENCE

    (By Research)

    DEPARTMENT OF MANAGEMENT STUDIES

    INDIAN INSTITUTE OF TECHNOLOGY MADRAS, INDIA

    FEB, 2016

    THESIS CERTIFICATE

  • 7/26/2019 Thesis 8-2-16.docx

    2/129

    This is to certify that the thesis titledOIL PRICE DETERMINANTS AND CO-MOVEMENT

    DYNAMICS, submitted by Srini!"!n.N, to the Indian Institute of Technology Madras,

    Chennai for the award of the degree of M!"#$r %& S'i$n'$ ()* R$"$!r'+, is a bonafide record of

    the research work done by him under my superision! The contents of this thesis, in full or in

    parts, hae not been submitted to any other Institute or "niersity for the award of any degree or

    diploma!

    Pr%&. M. T+$n%+i

    Research #uide

    $rofessor

    %ept! of Management &tudies $lace' Chennai

    IITMadras, ** *+ %ate'

    2

  • 7/26/2019 Thesis 8-2-16.docx

    3/129

    AC/NOLEGDMENTS

    I would like to epress my sincere gratitude to $rof! M! Thenmo-hi for guiding me throughout

    my research work! I would also like to thank my #eneral Test Committee members, %r! $!

    .rishna $rasanna and %r! "ma.anth %ash for their insightful comments! I would also like to

    thank $rof! T!/! .amalanabhan for his aluable suggestions! I would like to thank all the faculty

    of the %epartment of Management &tudies for helping me to gain knowledge through their

    teaching! I would like to thank all the staff of the %epartment of Management &tudies,

    0enkatraman &ir, Ra1endran &ir and 0asudean &ir for all their help and support! I would also

    like to thank Meenakshi Ma2am and 3-har for their help in the systems lab! I would like to thank

    &reeniasan and all the other staff of the %epartment!

    My #randmother, Mrs! 3!.! Thayaramma and my parents, 4ate Mrs! Reathi 5arasimhan and

    Mr! 0! 5arasimhan, and my brother Mr! 5! Raghuraman hae always supported me to pursue

    higher studies, and stood by me throughout and also helped me to go forward in academics! 5o

    amount of thanks would coney my gratitude for them!

    I would like to thank &hyaam, &hipra, 4akshmi akka, 3bhi1eeth &ir, 5arend &ir, &hashank,

    .ayal akka, 6ema for making my stay most memorable! I would like to thank my friends in the

    department &umeet, &halini, #iri sir, sharoon for their support!

    Srini!"!n.N

    ABSTRACT

    3

  • 7/26/2019 Thesis 8-2-16.docx

    4/129

    .eywords 7 Crude oil 7 &peculation 7 Marko Regime &witching Model 7 &tock market 7 8change rate 7 Co

    moement 7 9aelet 3nalysis 7 Continuous waelet transform

    This study inestigates the impact of fundamental, financial and speculatie factors on crude oil

    prices and analy-es the behaior of the arious determinants with respect to regimes! Thefundamental factors included in the study are :$8C production, :8C% stocks, :8C%

    consumption, :8C% 5et Imports, industrial production of China (in ;) and industrial production

    of India (in ;)! ** and tradeweighted

    "& dollar Inde and the speculatie component is captured using net long positions of non

    commercial traders! The study is conducted oer the period 3pril ?@@> to May A*?! Marko

    regimeswitching model is used analy-e the impact as it captures both non linearity and breaks!

    :ur empirical findings indicated that speculation affects the oil price positiely in higholatile

    state and has inerse effect in lowolatile state! 3t lowolatility regimes, fundamental, financial

    factors hae significant impact on the oil price, whereas at high olatility regimes, speculation

    has a significant effect on the oil price!

    The study further inestigated the relationship of oil price with echange rate and stock market

    by eamining the comoement between oil price and echange rate of nine oil importing

    countries stock market indices of fifteen ma1or oil importing countries! The differ from preious

    studies as the study eamines the macroeconomic dynamics of ma1or oilimporting countriesduring arious economic cycles across different freDuencies! The study includes fifteen ma1or oil

    importing countries oer the period A**+?! 9aelet Coherence analysis is used to etract the

    coherency! 8mpirical results indicate a high coherence between oil price and macroeconomic

    indicators across all the countries during the financial crisis! The echange rates hae negatie

    relationship with benchmark oil prices ecept for the echange rate of /apan in the long run and

    for the echange rate of &outh .orea in the medium run! &tock indices hae positie

    relationship with benchmark oil prices in both long and medium run! &=$ is leading the oil

    price, whereas &&8>*, 5ikkei AA>, 5I

  • 7/26/2019 Thesis 8-2-16.docx

    5/129

    macroeconomic indicators are obsered to change across freDuency and time! 8change rate

    offers diersification benefits, but stock market indices proide no diersification aenues since

    the pattern of comoement of stock market indices and oil prices are similar across all oil

    importing countries!

    The study has implications for regulators and policymakers! The study contributes to the

    literature in the field of oil price determination and coherency of oil price with macroeconomic

    indicators! The study used nominal price instead of real price to proide insight for traders and

    institutional inestors! The results aid the indiidual traders and institutional inestors in

    designing their portfolio for short, medium and long term time hori-ons!

    TABLE OF CONTENTS

    3C.5:948#%M85T&!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! iii

    5

  • 7/26/2019 Thesis 8-2-16.docx

    6/129

    3B&TR3CT!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! i

    4I&T :< T3B48&!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!iii

    4I&T :<

  • 7/26/2019 Thesis 8-2-16.docx

    7/129

    +!?! Introduction!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!+@

    +!A 4iterature ReiewJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!!!!!*

    +!+! %ata!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!?

    +! Conceptual flowJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ A

    +!> MethodologyJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ+

    +!>!?! Continous waelet transform!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!?

    +!A!A! 9aelet Coherence (9TC)!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!?

    +!! 9aelet coherence InferenceJJJJJJJJJJJJJJJJJJJJJJJJJ!! +

    +!H! 8mpirical ResultsJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ!!

    +!H!?! Relationship between benchmark oil price and echange rate of oil importing countries!! !

    +!H!A! Relationship between benchmark oil price and stock indices of oil importing countries! !!!@

    +!! &ummary!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!?

    CHAPTER CONCLUSION!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!@

    !? ?

    !?!>! :il price and stock indices' 9TCJJJJJJJJJJJJJJJJJJJJJJ!>A

    !A! Conclusions!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!@@

    !+! Contributions of the study!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!?*1

    !! Implications!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!?*?

    !>! 4imitations of the study!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!?*A

    R8

  • 7/26/2019 Thesis 8-2-16.docx

    8/129

    LIST OF TABLES

    T!)7$" Ti#7$ P!8$ N%.

    ?!? 8nergy Market &hare!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    ?!A $roed reseres of oil at the end of A*?+!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    ?!+ %ata %escription!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    ?! 8nergy rate %ata!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    ?!> &tock inde data!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    A!? &ummary of literature on determinants of crude oil!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    A!A %ata %escription!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    A!+ %escriptie statistics of fundamental, financial and speculatie ariables!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    A! B%& test results!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    A!> 8stimation results of Marko Regime &witching Regression ? !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    A! 8stimation results of Marko Regime &witching Regression A!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    A!H 8stimation results of Marko Regime &witching Regression +!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +!? Recent literature on relationship between the oil price and echange rate!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +!A Recent literature on relationship between the oil price and stock indices!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +!+ %escriptie statistics of echange rates of oil importing countries!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +! %escriptie statistics of stock indices of oil importing countries!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +!H Relationship during financial crisis!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    !? &ummary of hypothesis results!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    8

  • 7/26/2019 Thesis 8-2-16.docx

    9/129

    LIST OF FIGURESFi89r$" Ti#7$ P!8$ N%.

    ?!? Conditional means, 0olatilities and &moothed transition probabilities for model ? !!!!!!!!!!!!!!

    A! Conditional means, 0olatilities and &moothed transition probabilities for model A!!!!!!!!!!!!!!!

    A!H Conditional means, 0olatilities and &moothed transition probabilities for model + !!!!!!!!!!!!!!

    +!? Ma1or oil importing countries!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +!A 5ormali-ed oil price and echange rate of oil importing countries!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +!+ 5ormali-ed oil price and stock indices of oil importing countries !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +! Benchmark oil prices and echange rates 7 9aelet coherence (9TC)!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    +!> Benchmark oil prices and stock indices 7 waelet coherence (9TC)!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    9

  • 7/26/2019 Thesis 8-2-16.docx

    10/129

    ABBREVIATIONS

    inde

    5I 4G> inde (Indonesia)

    .:&$I .orea Composite &tock $rice Inde (&outh .orea)

    10

  • 7/26/2019 Thesis 8-2-16.docx

    11/129

    IB8F IB8F +> (&pain)

  • 7/26/2019 Thesis 8-2-16.docx

    12/129

    C+!:#$r 1

    In#r%;9'#i%n

    Crude oil constitutes a ma1or part in the world energy markets and supplies +* percent of the

    total world energy consumption and @* percent of the ehicular fuel consumption! &eeral

    countries in the world are either producers or consumers of crude oil! 6ence, it is the key drier

    of a nation2s economy and it also triggers economic cycles! Instability in oil price affects both

    deeloped as well as deeloping countries and olatility in oil price leads to potential

    ramifications oer a country2s economy and its financial markets!

    $reious research has focused more on the impact of demand and supply factors influencing oilprice! But oil is one of the highly traded commodities in the world with high participation from

    financial institutions in the deriatie segment and speculation in crude oil options and futures

    can play an actie role in establishing the price of crude oil! $recise estimation of the factors that

    affect the crude oil price helps in understanding the dynamics of crude oil price moements and

    helps in managing the risk pertaining to olatile oil prices! :il price series ehibit structural

    breaks and nonlinearity (Reboredo, A*?*) and may ehibit dynamic behaior with respect to the

    financial eents such as crisis, changes in goernment policy, changes in the business cycles and

    economic downturns! The economy becomes more difficult to manage when oil prices remain

    highly olatile, as higher olatility in crude oil prices, hae greater ramifications for different

    players in the economy and managing current account balance for goernments becomes a

    challenge! The determinants of crude oil price in higholatile period might be different from

    lowolatile period, and may differ during different economic phases! 6ence, an attempt is made

    in this study to eamine the effect of fundamental, financial and speculatie factors on crude oil

    prices during high and low olatile regimes!

  • 7/26/2019 Thesis 8-2-16.docx

    13/129

    company or indirectly by affecting the interest rate that is used to discount the future cash flows

    of a company! Change in stock prices and echange rates also impacts oil prices! But, the nature

    and etent of relationship between oil price and macroeconomic indicators may ary from time

    to time and understanding the pattern of relationship across time and freDuency hori-on becomes

    essential for traders and inestors! 6ence, we attempt to eamine the comoement between (i)

    oil price and stock inde and (ii) oil price and echange rate to capture the pattern of relationship

    across different time hori-on!

    E'%n%i' 8r%

  • 7/26/2019 Thesis 8-2-16.docx

    14/129

    The aboe table presents the ma1or fuel resources of the world! The annual increase in the supply of each fuel is

    presented! It also shows the current and forecasted share of each fuel in world energy market!

    9orld oil reseres as indicated in the B$ &tatistical Reiew of 9orld 8nergy (/une A*?) report

    is presented in Table ?!A! $roed reseres of oil are the estimated Duantities of oil that are

    obtained from geological and engineering information! These reseres are indicated with

    reasonable accuracy and can be recoered in the future from known reseroirs under eisting

    economic and operating conditions! $roed reseres in the world at the end ?@@+, A**+, and

    A*?+ is indicated in the table! It also presents the total Duantity in terms of both tones and barrels

    and the share of world reseres of ma1or oil producing nations!

    T!)7$ 1.2 Pr%$; R$"$r$" %& %i7 !# #+$ $n; %& 201

    $roed reseres of oil 8nd of A*?+

    9orldReseres

    ?@@+(Thousandmillionbarrels)

    A**+(Thousandmillionbarrels)

    A*?+(Thousandmillionbarrels)

    Thousandmilliontones

    thousandmillionBarrels

    share ofworld

    Reseres productionratio

    :8C% ?*! AH!> A@! +H!+ A! ?!H* ++!A

    5on:8C% @**! ?*! ?+H!H A**!@ ?+@!? >!+* >@!>

    :$8C HH!@ @?A!? ?A?+! ?H*!A ?A?!A H?!@* @*!+

    5on:$8CN A*!+ +A>!A +A! >*!? +?!@ A*!+* A

    8uropean "nionO !? ! *!@ ! *!* ?+

  • 7/26/2019 Thesis 8-2-16.docx

    15/129

    affects both deeloped as well as deeloping countries! 0olatility in the oil price leads to

    potential ramifications oer a country2s economy and its financial markets!

    8conomic recessions, inflation upsurge, and trade deficits are due to the large moements in oil

    price! $rice of crude oil in A**+ was at ;A@ a barrel! in A*??! The price was stable at ;?*> per barrel from A*?? to A*?! %r Oi7 Pri'$ S+%'="

    The ma1or eents that led to ma1or oil price shocks after the 9orld9arII include the &ue- Crisis

    of ?@>7?@>H, the :$8C oil embargo of ?@H+7?@H, the Iranian reolution of ?@H7?@H@, the

    Iran7IraD 9ar initiated in ?@*, the first $ersian #ulf 9ar in ?@@*@?, and the oil price spike of

    A**H7A** (6amilton, A*??)!

    In the initial postwar era, Teas Railroad Commission (TRC) played an important role in the

    world oil market! The permissible production leel was set by TRC based on the current market

    demand! 3s a result, the nominal oil price was constant from month to month! "nepected

    ariations in the oil price were witnessed during the etreme eents!

    S9$ Cri"i" (1?@61?@' The president of 8gypt, 5asser, has nationali-ed the &ue- Canal in

    ?@>! To counteract the 8gyptian president, Brittan and million barrels of oil per day was transported! It constitutes twothird of the total

    Brittan oil supplies! &upply disruption during this period led to sharp increase in oil price during

    the &ue- Crisis!

    OPEC E)!r8% (1?1?3rab members of :$8C etended their support to Israel against

    &yria and 8gypt by announcing an embargo on oil eports only to selected countries! 3s a result,

    15

  • 7/26/2019 Thesis 8-2-16.docx

    16/129

    :$8C has cut its total oil production which led to shortfall of world oil output by H!>! %ue to

    this the #ulf countries doubled the price of oil in ?@H! #eopolitical factors played a ma1or role

    in price spike during this period!

    Ir!n r$%79#i%n (1?1?? In order to fill the gap in the oil production due to 3rab :$8C

    member2s embargo, Iran increased its total oil production during the ?@H+7?@H!But, in ?@H,

    Iran was contended with large public protests which spread to the oil sector! The total oil

    production of Iran fell down by ! million barrels per day which constituted H of world

    production at the time, led to a price hike during ?@H7?@H@!

    Ir!nIr! !r (1?01?1 The oil supply from both the countries was lost during the war

    between Iran and IraD in ?@*! The total production loss accounted to about of world oil

    production at the time! 6oweer, this short fall in production was recoered from the other parts

    of the world with in a ery short span!

    T+$ 8r$!# :ri'$ '%77!:"$ (1?11?6 The oil demand of ma1or oil importing countries

    declined due to the price hike in ?@H*! The nominal oil price fell by A> by ?@*! &audi 3rabia

    has cut its threefourth of its production during ?@? and ?@>! But this did not improe the price

    of oil! In ?@, &audis lifted its production cuts! This led to a further collapse in oil price from

    ;AHbarrel in ?@> to ;?Abarrel in ?@! This was promising for oil importing countries but foroil eporting countries this was the ma1or shock!

    P$r"i!n G97& !r (1??01??1 In ?@@*, IraD inaded .uwait! The oil production from both

    countries accounted for @ of the total world oil production was lost during the war! The price of

    crude oil was doubled during this time! But, this spike sustained for ery short span!

    E!"# A"i!n Cri"i" (1??1?? In the mid of ?@@H, &outh .orea, Thailand, and some other

    3sian countries were sub1ected to financial stress! The price of oil soon followed their downfall

    since ma1or oil demand was from these countries! The price of oil has fallen below ;?A a barrel

    by the end of ?@@!

    G7%)!7 Fin!n'i!7 Cri"i" (200200 #lobal economy was booming during A** and A**>!

    #rowth in economy led to an increase in world oil consumption by > million barrels per day! 3s

    16

  • 7/26/2019 Thesis 8-2-16.docx

    17/129

    a result oil price increased rapidlyP the oil production slowly declined after A**>! The low

    production further added to the increase in the oil price during the global financial crisis!

    P79n8$ in %i7 :ri'$ (201 4ongterm trend in oil prices was determined by demand and supply

    factors! &hortterm moements are eplained using market sentiment and epectations!

    #eopolitical risks, supply disruptions, and production controls eercised by :$8C led to sharp

    decline of oil price from ;?*> per barrel! This decline was intensified due to policy change by

    :$8C! In A*?, due to geopolitical conflicts in some oil producing countries, :$8C announced

    its policy change which led to the appreciation of the "& dollar! Chen (A*?>) postulated that oil

    prices remain low in A*?> and will rise in A*? marginally! 4ongterm trend of oil production

    and consumption hae also played crucial roles in the recent decline in oil prices!

    1.2 D$#$rin!n#" %& 'r9;$ %i7

    The dynamics of oil price has receied much attention since the oil market witnessed rapid

    moements in its price! :il price from ;A@ a barrel in A**+ rose to ;?++ a barrel in A** and

    hae fallen again to ;+H a barrel in A*?>! &ome studies argue that the fluctuations in oil price

    were due to the structural transformations in oil market! :ther studies attributed this behaior to

    the seasonal and cyclical behaior of oil price! &teens (A**>) postulated that these different

    iews about the olatility in oil price reflect the diergent iews of the future change in oilprices!

    $reious studies on oil price studied basic correlations, causality relations, or economic linkages

    to oil price! But, these shortterm relationships cannot adeDuately capture the different shocks

    that influence the oil market! 3nd the longterm models are etremely sensitie to the underlying

    assumptions of the models!

  • 7/26/2019 Thesis 8-2-16.docx

    18/129

    sources of supply of crude oil are non:$8C nations and :$8C nations! There is a sudden

    increase in non:$8C production during A*?* crisis!

    The association between the fundamental factorsQdemand and supply and the oil price was

    approimated using a linear relationship in the most of the studies! But, nonlinear relationship

    eists between the determinants of crude oil price and the olatility of crude oil price! It is

    imperatie to accurately estimate the determinants of crude oil price using a nonlinear model!

    3ccording to the International 8nergy 3gency report A*?, the growth in the world demand for

    oil was approimately A per year oer the last two decadesP whereas the oil price fluctuation

    per year is way aboe A! This imbalance can be eplained using the factors other than

    fundamental factors! The other factors include speculatie factors, financial factors, and supply

    disruption factors! 6ence, the present study inestigates the fundamental, financial, and

    speculatie factors that eplain the olatility in crude oil price!

    1. In#$r-r$7!#i%n"+i: )$#

  • 7/26/2019 Thesis 8-2-16.docx

    19/129

    prices and emerging stock markets (Basher and &adorsky, A**P 6ammoudeh and 3leisa, A**P

    6ammoudeh and 6uimin, A**>)!

    .ilian and $ark (A**@) inestigated the relationship between the "& stock prices and crude oil

    price shocks for the period ?@H+7A**! They obsered asymmetric effect of oil price shocks on

    stock market! They found that oil supply shocks hae much less impact on stock prices but oil

    demand shocks depressed stock prices! 3pergis and Miller (A**@) analy-ed the effect of oil price

    shocks on the stock prices using a &03R approach! They considered the monthly data of eight

    deeloped economies (3ustralia, Canada,

  • 7/26/2019 Thesis 8-2-16.docx

    20/129

    Cifarelli and $aladino (A*?*) used a multiariate generali-ed autoregressie conditional

    heteroskedasticity inmean (#3RC6M) model to estimate the effect of oil price shifts on

    echange rate! 3kram (A**@) suggested that a weaker "&% led to higher oil prices and that "&%

    fluctuations accounted for olatility in oil price! Most of the empirical literature on oil and

    echange rate dependency proides eidence of oil7echange rate comoement (Reboredo,

    A*?A) only in time domain! It is also important to understand the relationship between oil price

    and echange rate in freDuency domain as it indicates short and long term dynamics! But ery

    little literature eists on oil7echange rate dependency in both time and freDuency domains!

    :il prices and echange rates comoements hae implications for a central bank that aims to

    achiee a desired leel of appreciationdepreciation for its currency against foreign currencies!

    Comprehending the relation between oil and echange rate markets is crucial as it impacts manykey financial and economic aspects of a country! The impact of oil price shock is significant

    mainly for oilimporting countries! 6ence, understanding the correlation between oil prices,

    echange rates, and stock market is important!

    1. N$$; &%r #+$ S#9;*

  • 7/26/2019 Thesis 8-2-16.docx

    21/129

    consumption alone is not appropriate, since it ecludes ma1or oil consuming countries like China

    and India! 6ence, use of proy ariables to account for the increasing consumption from

    emerging economies such as China and India is pertinent to capture the global consumption

    effectiely!

    Contemporary research in the field of oil price dynamics indicates that financial factors also

    affect oil price apart from fundamental factors! Many studies hae shown significant impact of

    echange rate on crude oil prices (3mano and 0an 5orden, ?@@P Basher et al!, A*?AP Beckmann

    and C-uda1, A*?+), but most of them hae used 588R (Breitenfellner et al!, A**@), R88R

    (:riawote and 8riemo, A*?AP Breitenfellner et al!, A**@) and "& dollar echange rate, which is

    based on the transaction between "& and its trading partners! 8amining the impact of effectie

    echange rate of dollar inde has limitations in accounting for the global trading of oil in dollars!

    6oweer, using a comprehensie measure such as %ollar inde (broad) would capture the

    complete dynamics of oil trading! 6ence, it is pertinent to eamine the impact of dollar inde

    (broad) on oil price than looking at the impact of effectie echange rate on oil prices!

    $rior research hae also found significant impact of stock inde on crude oil prices (Basher et al!,

    A*?AP Cifarelli and $aladino, A*?*P Breitenfellner et al!, A**@)! 9ith the adancement of

    fianancialisation, forward and futures market impact oil prices! If the epected future oil spot

    price is greater than the futures price, there will be a premium to etract oil from well! But, the

    impact of spread between the current spot price and a year ahead futures price of oil has not been

    considered by prior studies! Considering ?Amonth basis which measures the premium of

    holding storing a crude oil rather than a deriatie product may proide more insight on crude

    oil price discoery!

    &ornette et al!, (A**@) postulated that the oil price shocks during crisis are due to speculatie

    factors and speculatie trading can influence the oil prices without any change in fundamental

    factors (6amilton, A**@)! &peculation ariable such as feedback trading was found to haenegatie effect on oil prices (Cifarelli and $aladino, A*?*)! 9hile, noncommercial net long

    positions (futures) are found to hae positie effect oil price (

  • 7/26/2019 Thesis 8-2-16.docx

    22/129

    &peculatie actiity based on noncommercial net long positions (futures and options) may

    capture the impact of speculation on oil price more effectiely!

    :il price series ehibit structural breaks and nonlinearity (Reboredo, A*?*) and may ehibit

    dynamic behaior with respect to the financial eents such as crisis, changes in goernment

    policy, changes in business cycles and economic downturns! 3s the behaior of the oil price

    determinants change oer time, understating the oil price and its determinants behaior with

    respectie regimes becomes crucial! "sing ordinary regression model fails to identify the

    dynamic linkage between oil prices and its determinants across regimes! Most of the eisting

    literatures eamine the relationship using traditional linear time series models such as 03R

    08CM, cointegration and structural 03R! 6oweer, oil price is found to hae structural breaks

    and may not be linear and most of the studies eamine casual long term relationship of crude oil

    with only one or few ariables! 4inear models could suffer from possible misspecification or

    omitted ariable bias and the relationship between oil price and determinants may ary oer

    time! Incorporating or using the model, which accounts for structural break and capturing the

    effect at different periods would help in accurate estimation of oil price! 0ery few studies hae

    used nonlinear models such as CCC #3RC6M and Bayesian Model aeraging (Breitenfellner

    et al!, A**@)! 6oweer, these models cannot estimate the relationship with respect to different

    market phases and fail to address structural breaks in the data! Marko Regimeswitching model

    proides a fleible framework to model structural breaks, dynamic shifts and dynamic

    relationships! Markoregime switching methodology is largely used for finding nonlinear

    causality (

  • 7/26/2019 Thesis 8-2-16.docx

    23/129

    with the appreciating echange rate (3mano and 0an 5orden, ?@@P BenassyGuere et al!, A**HP

    5arayan et al!, A**P Basher et al!, A*?AP Beckmann and C-uda1, A*?+)! But other studies argued

    that increase in oil prices is associated with the depreciating echange rate (9ang and 9u,

    A*?A)! 9hile bidirectional causality eists between oil price and echange rates after crisis

    (%ing = 0o, A*?A), at large time hori-ons (Benhmad, A*?A) and at higher time scales (Tiwari et

    al!, A*?+), Iwayemi and

  • 7/26/2019 Thesis 8-2-16.docx

    24/129

    perspectie considering nominal prices! 6ence, there is a need to focus on traders and

    institutional inestor2s perspectie and eamine the comoement between benchmark oil price

    with (i) nominal echange rates and (ii) stock indices of ma1or oilimporting countries using

    waelet coherence approach!

    1.@ O)>$'#i$" %& #+$ S#9;*

    The purpose of this study is to eamine the factors that determine the crude oil price and the

    comoement dynamics with macroeconomic factors! It is important to understand the factors

    affecting crude oil price to accurately model crude oil price behaior with respect to the changes

    in the economic cycles and olatile regimes! It is also pertinent to understand the impact of crude

    oil price on the economy and eplore the link between oil prices and arious macroeconomic

    ariables for a large set of oilimporting countries! 6ence, the specific ob1ecties of the study

    are

    ?! to eamine the impact of fundamental, financial, and speculatie factors on 9TI crude oil

    prices at high and low olatile regimes!

    A! to eamine the comoement between oil price and echange rate of oil importing

    countries across different time and freDuency hori-ons!

    +! to eamine the eistence of comoement between oil price and &tock Indices of oil

    importing countries across different time and freDuency hori-ons!

    In this study, comoement between oil price and echange rates, and oil price and &tock Indices

    is perceied from perspectie of both the traders and institutional inestors! Traders and

    institutional inestors hae aried inestment hori-ons, as they hae aried risk profiles!

    1.6 H*:%#+$"i"

    6ypothesis for the study has been formulated and tested based on the gap from the eisting

    literature! The hypothesis is as follows

    6?' $resence of linear relationship

    24

  • 7/26/2019 Thesis 8-2-16.docx

    25/129

    6A' There is no significant impact of speculatie factors on 9TI oil prices

    6+' There is significant impact of speculatie factor on 9TI oil prices controlling for financial

    factors

    6' There is no significant impact of speculatie factors on 9TI oil prices controlling for

    fundamental, financial, and supply disruption factors

    6>' The effect of 9TI crude oil prices on echange rate of oil importing countries is not

    statistically significant

    6' The effect of 9TI crude oil prices on &tock Inde of oil importing countries is not

    statistically significant

    Fi89r$ 1.1 Fr!$**

    ?A month B3&I& (

  • 7/26/2019 Thesis 8-2-16.docx

    26/129

    1. D!#! !n; S!:7$

    The sample used for the study comprises of monthly 9TI crude oil spot price traded at 5ew

    Eork Mercantile 8change (5EM8F)! The 9TI crude oil data is collected from 8nergy

    Information 3gency! The ariables effecting crude oil price are grouped into four broad

    categories, namely' (i) demand factors, (ii) supply factors, (iii) financial factors, and (i)

    speculatie factors! The demand factors include and for the ariables such as :8C%

    consumption, :8C% imports, China Industrial $roduction, and India Industrial $roduction! The

    supply factors include Total oil Rigs, :$8C production, and :8C% inentory! **, Basis 7 :il ( 7 *>A*? 8nergy Information 3dministration

    26

  • 7/26/2019 Thesis 8-2-16.docx

    27/129

    Imports (:8C%) *?@@> 7 *>A*? 8nergy Information 3dministration

    C6I53 Industrial Production (US$) *?@@> 7 *>A*? 9orld bank

    I5%I3 Industrial Production (US$) *?@@> 7 *>A*? 9orld bank

    S9::7* V!ri!)7$"

    :$8C $roduction *?@@> 7 *>A*? 8nergy Information 3dministration

    :8C% Inentory *?@@> 7 *>A*? 8nergy Information 3dministration

    Financialization - Variables

    %ollar Inde *?@@> 7 *>A*? &t! 4ouis A*? 8nergy Information 3dministration

    Speculation - Variables

    5et 4ong $ositions (5onCommercial) 7 :il( 7 *>A*? Commodity

  • 7/26/2019 Thesis 8-2-16.docx

    28/129

    T!)7$ 1. E'+!n8$ r!#$ D!#!

    E'+!n8$ C%9n#r* C9rr$n'* N!$

    "&%C5E China Chinese Euan

    "&%8"R 8uro Region 8uro

    "&%I%R Indonesia Indonesian Rupiah

    "&%I5R India Indian Rupee

    "&%/$E /apan /apanese Een

    "&%.R9 &outh .orea &outh .orean 9on

    "&%% &ingapore &ingapore dollar

    "&%TRE Turkey Turkish 4ira

    "&%T9% Taiwan 5ew Taiwan %ollar

    Relationship between oil price and echange rate is etensiely studied (3mano and 0an

    5orden, ?@@P Chen and Chen, A**HP Basher et al!, A*?AP Beckmann and C-uda1, A*?+)!

    C+in$"$ Y9!n (CNY is the currency for china! Chinese Euan is pegged with "& %ollar! It

    follows a fied echange rate system! 6uang and #uo (A**H) studied the impact of oil price

    shock on the trend moements of ChinaLs real echange rate!

    E9r% (EUR is the currency for the following countries 7 3ustria, Belgium4uembourg,

    Cyprus,

  • 7/26/2019 Thesis 8-2-16.docx

    29/129

    In;i!n R9:$$ (INRis the official currency of India! Indian currency is a managed float! RBI

    trades freDuently on "&%I5R, thus it is a de facto controlled echange rate! :ther rates (such as

    the 8"RI5R and I5R/$E) hae the olatility typical of floating echange rates! #hosh (A*??)

    studied relationship between the crude oil price and "&%I5R for the period starting from And

    /uly A**H to Ath %ecember A**!

    4!:!n$"$ Y$n (4PYis the currency for /apan! The country adopted a floating or fluctuating

    echange rate system! 5aka1ima and 6amori (A*?A) studied the casualty between crude oil and

    /$E"&% echange rate! The data period considered for the study spans from ?st &eptember

    A**> to A@th /uly A*??!

    S%9#+ /%r$!n %n (/R is the currency of &outh .orea! They hae allowed freefloat

    echange rate from ?@@H! Masih et al! (A*??) studied the implications of oil price fluctuations

    and oil price olatility on echange rate of &outh .orea!

    Sin8!:%r$ D%77!r" (SGDis the currency of &ingapore! It follows a managed float echange rate

    system!

    T9r=i"+ Lir! (TRYis the official currency of Turkey and the Turkish Republic of 5orthern

    Cyprus! It follows freefloat echange rate system! &oytas (A*??) eamined the longrun and

    shortrun transmissions of information between the world oil price, Turkish interest rate, Turkish

    lira7"& dollar echange rate, and domestic spot gold and siler price!

    N$< T!i

    They adopted freefloat echange rate system since ?@@! RafiD et al! (A**@) empirically

    eamined the impact of oil price olatility on key macroeconomic indicators of Thailand!

    The sample stock indices that were considered for the study are as follows! 5ikkei AA> inde

    (/apan 5.E), 5I* (India 5I inde (Indonesia 4G>), .orea Composite &tock

    $rice Inde (&outh .orea .:&$I), IB8F +> (&pain IB8F),

  • 7/26/2019 Thesis 8-2-16.docx

    30/129

    includes from th /anuary A*** to +*th %ecember A*? for all the indices ecept for &&8 >*!

    &ample period for &tock inde of china (&&8>*) starts from >th /anuary A** to +*th %ecember

    A*?! The echange rate and stock inde data is collected from the Bloomberg database! Table

    ?!> summari-es the stock inde data!

    T!)7$ 1.@ S#%'= In;$ ;!#!

    In;$ C%9n#r* In;$ n!$

    5.E /apan 5ikkei AA>

    5I Indonesia 4G >

    .:&$I .orea .orea Composite &tock $rice Inde

    IB8F &pain IB8F +>

    and the base alue was set at ?***! It is a freefloat market capitali-ationweighted inde!

    30

  • 7/26/2019 Thesis 8-2-16.docx

    31/129

    S#!n;!r; P%%r" @00 is an 3merican &tock Market inde! It comprises of >** large

    companies (based on their market capitali-ation), that are listed on either 5ew Eork &tock

    echange (5E&8) or 5ational 3ssociation of &ecurities %ealers 3utomated Guotations

    (53&%3G)! It is a freefloat capitali-ationweighted inde! The inde was in eistence from

    ?@>H! The components of the &=$ >** are selected based on eight parameters! They are market

    capitali-ation, liDuidity, domicile, public float, sector classification, financial iability, length of

    time publicly traded, and listing echange!

    L @ in;$ is Indonesian inde! It was launched in most

    liDuid companies listed on the Indonesia &tock 8change! It is a market capitali-ationweighted

    inde! ?+ /uly ?@@ was used as the base day, with an inde base alue of ?**! It consists of

    ma1or sectors such as agriculture, finance, mining, infrastructure, etc! It captures H* of thedomestic market capitali-ation!

    Ni==$i 22@ in;$is a stock market inde for /apan! It comprises of AA> top companies that are

    traded on the Tokyo &tock 8change (T&8)! The components of the inde are reiewed once a

    year! It is a priceweighted inde! The 5ikkei AA> &tarted on H &eptember ?@>*!

    /%r$! C%:%"i#$ S#%'= Pri'$ 200 In;$ (/OSPI 200 is the inde of &outh .orea! It

    comprises of A** large publiclytraded companies on the .orean &tock 8change! This is marketcapitali-ationbased weighted inde! .:&$I was introduced in ?@@*! The base alue of the inde

    was ?**! It has oer H* market alue of the composite stock price inde which includes all the

    stocks that are traded on the .orean &tock 8change, and so moes along with the .:&$I inde!

    IBE3 @is the benchmark stock market inde of &pain! The inde comprises of +> most liDuid

    &panish stocks that are traded on Madrid &tock 8change #eneral Inde! It was initiated in ?@@A!

    It is a market capitali-ationweighted inde! The inde is reiewed twice annually!

    FTSE MIB 0 is the primary bench mark stock market inde of Italy! The inde consists of the

    * top liDuid stocks that are traded on the Borsa Italiana, the Italian national stock echange! It

    represents the large cap component of the

  • 7/26/2019 Thesis 8-2-16.docx

    32/129

    S#r!i#" Ti$" In;$ (STIis the benchmark inde for the &ingapore stock market! It is market

    capitali-ationweighted stock market inde! It constitutes the top +* liDuid companies that are

    listed and traded on the &ingapore echange! It is 1ointly calculated by &ingapore $ress 6oldings

    (&$6), &ingapore 8change (F), and

    of the total market capitali-ation

    D$9#"'+$r A=#i$nin;$ (DA3 0 is the benchmark inde of #ermany! It consists of +* ma1or

    #erman companies trading on the

  • 7/26/2019 Thesis 8-2-16.docx

    33/129

  • 7/26/2019 Thesis 8-2-16.docx

    34/129

    S#%'= !r=$# In;$ (SP @00is the proy for the growth of the economy and is an alternate

    inestment ehicle (Cifarelli and $aladino, A*?*P Breitenfellner et al!, A**@)

    Tr!;$-

  • 7/26/2019 Thesis 8-2-16.docx

    35/129

    WTI oil=Intercept+1NC NET LONG POS+

    2SP500+

    3Dollar Index+

    412M Basis+

    9TI crude oil price is used as dependent ariable! 5C 58T 4:5# $:& represents non

    commercial trader2s net long position, &=$>** represents the &=$ >** inde, %ollar inde

    represents the Tradeweighted "& dollar inde (Broad) and ?AM Basis represents the ?Amonth

    basis between 9TI crude oil

  • 7/26/2019 Thesis 8-2-16.docx

    36/129

  • 7/26/2019 Thesis 8-2-16.docx

    37/129

    D$#$rin!n#" %& %i7 :ri'$ A M!r=%-r$8i$ "

  • 7/26/2019 Thesis 8-2-16.docx

    38/129

    (iii) The impact of fundamental, financial and speculation factors on 9TI oil price at high

    and low olatility regimes!

    :il price series ehibit structural breaks and nonlinearity (Reboredo, A*?*) and may ehibit

    dynamic behaior with respect to the financial eents such as crisis, changes in goernment

    policy, changes in the business cycles and economic down turns! 3s the behaior of the oil price

    determinants2 changes oer time, understating the oil price and its determinants2 behaior with

    respectie regimes becomes crucial! "sing ordinary regression models fail to identify the

    dynamic linkage between oil prices and its determinants across regimes! Marko Regime

    switching model proides a fleible framework to model structural breaks, dynamic shifts and

    dynamic relationships! 6ence, Marko Regimeswitching model was used to eamine the

    determinants of 9TI oil prices at high and low olatile regimes!

    8mpirical results of our study indicate that speculatie factor plays an important role only in high

    olatile state! 9hereas in low olatile state fundamental, and financial factors a ma1or role in

    determining the oil price! The study proides a holistic iew about price formation in the crude

    oil market by incorporating determinants from fundamental, financial and speculatie factors!

    The study has implications for policymakers, financial institutions and inestors, as we indicate

    the determinants that affect the crude oil price for each regime separately!

    The remainder of this chapter is organi-ed as follows' &ection A!? describes a brief literature on

    determinants of crude oil, elaborates research gap and motiation! &ection A!+ discusses the

    framework of the study! &ection A! presents the data! &ection A!> eplains the Methodology!

    &ection A! discusses the empirical results and &ection A!H concludes the chapter!

    2.2 Li#$r!#9r$ %n &!'#%r" in&79$n'in8 Cr9;$ Oi7

    The price fluctuations in the crude oil market are either due to changes in fundamental factors or

    financial factors or due to speculation actiities! $reious literature analy-ed the crude oil priceto determine the factors affecting the crude oil price (Ee et al!, A**AP Breitenfellner et al!, A**@P

    Isabel 0ansteenkiste, A*??)! But, the literature is inconclusie in demonstrating what dries the

    crude oil prices!

    38

  • 7/26/2019 Thesis 8-2-16.docx

    39/129

    Ee et al! (A**A) forecasted the 9TI crude oil spot price considering :8C% petroleum inentory

    leels as independent ariable! They hae used monthly data of the oil price and petroleum

    inentory oer the period starting from /anuary ?@@A to **, 588R, R88R, 4agged oil price, and ariables pertaining to

    :$8C structure! Bayesian model aeraging techniDue is used for their study! The data period

    considered spans from ?@> to A**! 6e found that in ?@@*s and in A***s, fundamental factors

    played a significant role in driing the oil prices!

    4in and Tamakis (A*?*) applied eent study methodology to study the effect of :$8C

    announcements on the 9orld :il prices! The study used daily data of different crude oil

    benchmarks such as 3laska 5orth &lope, Brent blend, Bonny 4ight, %ubai

  • 7/26/2019 Thesis 8-2-16.docx

    40/129

    :$8C announcements hae significant effect on the oil prices! But the magnitude of the effect is

    based on the preious price regime! They also found that Guota cuts lead to positie price

    returns! Guota increases lead to negatie price returns in weak and normal price regimes! 5o

    change in Guota leads to insignificant or negatie price returns!

    Cifarelli and $aladino (A*?*) inestigated the effect of speculation on oil price fluctuations! They

    hae considered the weekly prices of the 9TI :il &pot,

  • 7/26/2019 Thesis 8-2-16.docx

    41/129

    between the oil and echange rate! 5o Cointegrating relationship is obsered among these

    factors! The :il and echange rate had Inerse relationshipP rise in the oil price depreciates the

    %ollar echange rate against Taiwan currency!

    Basher et al!, (A*?A) studied the dynamic relationship between the oil price, echange rate, and

    emerging market stock prices during the period /anuary ?@ to %ecember A**! They hae

    considered supply, demand and financial ariables for the study! They hae estimated the

    relationship using &tructural 03R! 5o Cointegration has been found among the ariables! They

    also obsered that the rise in the oil price tend to depress emerging market stock prices! They

    noted that positie oil price shock lead to an immediate drop in the tradeweighted echange rate

    and depreciation of currencies of oil importers in the long run!

    Miller and Ratti (A**@) analy-ed the longrun relationship between the world price of crude oil

    and stock markets oer the period starting from /anuary ?@H? to March A**! 0ariables used in

    the study are stock indices, interest rates, Industrial production, and 9orld crude oil price! 0ector

    8rror correction model is used to study the relationship! Their empirical results confirmed the

    eistence of longrun relationship among the ariables and structural Breaks They also found that

    the increase in oil price has negatie effect on stock Indices!

    Cheillon and Rifflart (A**@) analy-ed the real rice of crude oil focusing on physical market!

    Their data spans from ?st Duarter ?@@ to th Duarter A**>! 03R methodology is applied to

    analy-e the real oil price using supply and demand ariables! They found that the epected

    demand coerage ratio, 5on:8C% &tocks has negatie effect on oil price! 9hereas, :8C% oil

    demand, 5on:8C% oil demand has positie effect on oil price! It is also found that Inasion of

    .uwait has positie effect whereas the Inasion of IraD has negatie effect on oil price! They

    also found that Cointegrating ectors of :$8C Duotas and price and target has negatie effect

    oil price whereas the Cointegrating ectors of :8C% stocks and demand has positie effect on

    the oil price

    .aufmann and "llman (A**@) eamined the causal relationships between prices of the crude oil

    in arious markets! The sample data of their study consists of spot oil data starts fromAnd

    /anuary ?@ for 9TI and %ubai

  • 7/26/2019 Thesis 8-2-16.docx

    42/129

    9TI, +rd 3pril A** %ubai

  • 7/26/2019 Thesis 8-2-16.docx

    43/129

  • 7/26/2019 Thesis 8-2-16.docx

    44/129

  • 7/26/2019 Thesis 8-2-16.docx

    45/129

    $rior research hae also found significant impact of stock inde on crude oil prices (Basher et

    al.,A*?AP Cifarelli and $aladino, A*?*P Breitenfellner et al., A**@)! 9ith the adancement of

    financiali-ation, forward and futures market impact oil prices! If the epected future oil spot

    price is greater than the futures price, there will be a premium to etract oil from well! But, the

    impact of spread between the current spot price and a year ahead futures price of oil has not been

    considered by prior studies! Considering ?Amonth basis which measures the premium of

    holding storing a crude oil rather than a deriatie product may proide more insight on crude

    oil price discoery!

    &ornette et al!, (A**@) postulated that the oil price shocks during crisis are due to speculatie

    factors and speculatie trading can influence the oil prices without any change in fundamental

    factors (6amilton, A**)! &peculation ariable such as feedback trading was found to hae

    negatie effect on oil prices (Cifarelli and $aladino, A*?*)! 9hile, noncommercial net long

    positions (futures) are found to hae positie effect oil price (

  • 7/26/2019 Thesis 8-2-16.docx

    46/129

  • 7/26/2019 Thesis 8-2-16.docx

    47/129

    To ascertain the effect of fundamental factors, we hae used the ariables such as :$8C

    production, :8C% stocks, :8C% 5et Imports! The ma1or demand for oil comes from emerging

    economies and :C8% countries! To estimate the demand from :8C% countries we consider

    :8C% consumption! &ince the consumption data for the emerging economies is noneisting,

    Industrial production of China (in ;) and industrial production of India (in ;) is considered as

    proy for the demand from emerging countries! 6oweer, ariables such as industrial production

    of China (in ;) and industrial production of India (in ;) were not used in the literature! 9e find

    the relatie effect of the ariables across periods of low and high olatility!

    8isting literature used financial ariables such as &=$ >** (Cifarelli and $aladino, A*?*), Real

    emerging eDuity prices (Basher et al!, A*?A), %ollar Inde (Basher et al!, A*?A), 588R, R88R

    (Breitenfellner et al!, A**@), 4agged oil price (Breitenfellner et al!, A**@), :$8C implicit price

    target (Cheillon and Rifflart, A**@)! In our study, we hae included financial ariables such as

    &=$ >**, %ollar Inde and a ?Amonth basis! In order to capture the speculatie effect we hae

    included noncommercial net long position (both futures and options), as eisting net long

    position influences the price moements of the security!

    Most of the eisting literature eamined this relationship ma1orly using traditional linear time

    series models such as 03R 08CM, cointegration and structural 03R! :il price is also found to

    hae Breaks and longrun relationship in the ariables (Miller and Ratti, A**@)! 6ence these

    linear models suffer from possible misspecification or omitted ariable bias! Moreoer, the

    relationship between oil price and other determinants may ary oer time! Incorporating or using

    the model, which accounts for structural break and the effect could change at different periods

    would help in accurate estimation of oil price!

    0ery few studies hae used nonlinear models such as CCC #3RC6M and Bayesian Model

    aeraging! 6oweer, these models cannot estimate the relationship with respect to different

    market phasesP also, they fail to address breaks! Markoregime switching methodology islargely used for finding nonlinear causality (

  • 7/26/2019 Thesis 8-2-16.docx

    48/129

    T!)7$ 2.1 S9!r* %& 7i#$r!#9r$ %n ;$#$rin!n#" %& #+$ 'r9;$ %i7

    3uthors &ample :b1ectie %eterminants Model Results

    .ilian and4ee (A*?)

  • 7/26/2019 Thesis 8-2-16.docx

    49/129

    Basher et al!,(A*?A)

    /anuary?@7%ecemberA**

    Relationship between oilprice, echange rate, andemerging market stockprices

  • 7/26/2019 Thesis 8-2-16.docx

    50/129

    Cheillonand Rifflart(A**@)

    ?G ?@@7GA**>

    3naly-e the real rice ofcrude oil with focus onphysical market

    ! Inasion of IraD (V! Cointegrating ec

    of :$8C Duotas anprice and target (V)

    H! Cointegrating ecof :8C% stocks andemand(U)

    .aufmannand "llman(A**@)

    &pot andfutures data

    8amine causalrelationships betweenprices of crude oil inarious market

    5ear month and far mothfutures for 9TI, Brent,%ubai, spot price of 9TI,Brent, Bonny, Maya and%ubai

    Modified Test for CoIntegration(#rangerand 4ee, ?@@)

    Cointegration eists betwall the crude oil benchmar

    .ilian (A**@)/anuary ?@H+to %ecemberA**H

    Identify %emand andsupply shocks foreplaining fluctuationin the real oil price

  • 7/26/2019 Thesis 8-2-16.docx

    51/129

    capacity, oil reseres, oilsupply, oil stocks, capacityutili-ation, eplorationcost, Baltic dry inde,dummy (hurricane, war),:$8C reseres, :$8CDuota, &audi Duota and oilsupply (:$8C share)

    :ther factors'588R, R88R,&=$, 5etpositions, and 4agged oilprice!

    3kram(A**@)

    /anuary ?@@*7 3pril A**H

    8amine therelationship amongcommodity prices,interest rates and thedollar

  • 7/26/2019 Thesis 8-2-16.docx

    52/129

    The hypothesis of our study is as follows'

    a! 5ull hypothesis for the presence linear relationship

    H0a: Presence o linear relations!i"

    b! 5ull hypothesis for the impact of speculatie factors on the 9TI oil prices

    H0b: T!e coeicient o t!e s"eculati#e #ariable is ero

    c! 5ull hypothesis for the impact of speculatie factor on the 9TI oil prices controlling for

    financial factors

    H0c: T!e coeicient o t!e s"eculati#e #ariable is ero controllin% or inancial actors

    d! 5ull hypothesis for the impact of speculatie factors on the 9TI oil prices controlling for

  • 7/26/2019 Thesis 8-2-16.docx

    53/129

    The data is collected from databases such as 8nergy Information 3gency, 9orld Bank, &t! 4ouis

  • 7/26/2019 Thesis 8-2-16.docx

    54/129

    Cheillon and Rifflart, A**@P Ee et al!, A**A)! 6igh Inentory is epected to hae negatie effect

    on the oil prices in the short run (V)!

    OECD C%n"9:#i%n !n; I:%r#", measure the actual demand from deeloped economies and

    from emerging economies! In;9"#ri!7 :r%;9'#i%n %& C+in! !n; In;9"#ri!7 :r%;9'#i%n %&

    In;i!are included in the study, as they are the two ma1or oil consuming and importing countries

    outside the :8C% region! The demand factors are epected to hae a positie effect onthe oil

    price (U)!

  • 7/26/2019 Thesis 8-2-16.docx

    55/129

    negatie net long positions is epected to decrease the oil price (V)!Table A!A summari-es the

    ariables that are used, sample period considered for each ariable and source it is taken from!

    T!)7$ 2.2 D!#! D$"'ri:#i%n

    DETERMINANTS SAMPLE PERIOD SOURCE

    Oi7 B$n'+!r=

    4agged 9TI &$:T (5EM8F) *?@@> 7 *>A*? 8nergy Information 3dministration

    Fundamental Variables

    D$!n; V!ri!)7$"

    Consumption (:8C%) *?@@> 7 *>A*? 8nergy Information 3dministration

    Imports (:8C%) *?@@> 7 *>A*? 8nergy Information 3dministration

    C6I53 Industrial Production (US$) *?@@> 7 *>A*? 9orld bank

    I5%I3 Industrial Production (US$) *?@@> 7 *>A*? 9orld bank

    S9::7* V!ri!)7$"

    :$8C $roduction *?@@> 7 *>A*? 8nergy Information 3dministration

    :8C% Inentory &tock *?@@> 7 *>A*? 8nergy Information 3dministration

    Financialization variables

    %ollar Inde (Broad) *?@@> 7 *>A*? &t! 4ouis A*? 8nergy Information 3dministration

    Speculation variables

    5et 4ong $ositions (5onCommercial) 7 :il( 7 *>A*? Commodity

  • 7/26/2019 Thesis 8-2-16.docx

    56/129

    Fi8. 2.1 I:!'# %& S:$'97!#i%n %n TI Cr9;$ %i7 :ri'$ 7%< !n; +i8+ !ri!n'$ "#!#$"

    Fi8. 2.2 I:!'# %& S:$'97!#i%n %n TI Cr9;$ %i7 :ri'$ "$ri$" '%n#r%77in8 &%r &in!n'i!7

    !ri!)7$" 7%< !n; +i8+ !ri!n'$ "#!#$"

    56

    9TI Crude oil price

    (&pot 7 5EM8F)

    SPECULATION

    5onCommercial net 4ong positions (

  • 7/26/2019 Thesis 8-2-16.docx

    57/129

    Fi8. 2. I:!'# %& &9n;!$n#!7", &in!n'i!7 !n; ":$'97!#i$ &!'#%r" %n TI 'r9;$ %i7 :ri'$

    7%< !n; +i8+ !ri!n'$ "#!#$"

    57

    F9n;!$n#!7

    Oi7 B$n'+!r=

    9TI &pot price (5EM8F)

    DEMAND

    :8C% Imports

    :8C% Consumption

    Industrial production India

    Industrial production 7 China

    SUPPLY

    :$8C $roduction

    :8C% &tocks

    9TI Crude oil price(&pot 7 5EM8F)

    SPECULATON

    5onCommercial net 4ong positions

    (

  • 7/26/2019 Thesis 8-2-16.docx

    58/129

  • 7/26/2019 Thesis 8-2-16.docx

    59/129

    &eeral studies in the literature hae used Markoregime switching models to inestigate

    nonlinearity and asymmetry (6amilton, ?@@P #ray, ?@@P .rol-ig, ?@@P .rol-ig, A***P 3rtis et

    al!, A**)! 6amilton (?@@) proposed that Regime shifts in mean and ariance in a time series

    ariable ty

    can be modeled using Markoregime switching autoregressie model (M&3R) of

    "thorder as follows

    ?

    ?

    [( ) U ( [( )) ( ) !"

    t t i t t i t t

    i

    y s y s s =

    = +

    (A!A)

    9here \iis the slope coefficient of the autoregressie term (y t V ?)! [ and ] are the mean and

    standard deiation based on the state at time t(st! ytrepresents the 9TI crude oil spot price, this

    M&3R methodology allows us to model the potential regime shifts in the crude oil spot price!

    .rol-ig (?@@) deeloped M&03R, which is a generali-ation of the M&3R model (6amilton,

    ?@@) and used to find the causal relationship between endogenous ariables! &eeral eisting

    literature used M&03R for this purpose (.rol-ig, ?@@HP 9arne, A***P 8hrmann et al!, A**+P

    $saradakis et al!, A**>)! In these models, the intercept term, autoregressie slope coefficient and

    the error ariance are regime dependent! The M&03R is modeled as follows'

    ? ?

    ? A + ,

    ? ?

    ( ) ( ) ( ) ,t ) t t * ) t t * t r t * *

    r s r s + # s u = =

    = + + + (A!+)

    ? ?

    B > ,

    ? ?

    ( ) ( ) ( ) !t ) t t * ) t t * t + t * *

    + s + s r # s u = =

    = + + + (A!)

    4et2s assume rtand+tare two endogenous ariables! utis the innoation process, with ariance

    #(st) based on the statestat time t, follows an irreducible twostate Marko process defined by

    the transition probabilities (Pi) between states as follows'

    ( ?)Y Z,i) t t P " s ) s i= = = 9ith

    A

    ?

    ?i))

    P=

    =for all i, ^?, A_,

    9here,

    59

  • 7/26/2019 Thesis 8-2-16.docx

    60/129

  • 7/26/2019 Thesis 8-2-16.docx

    61/129

  • 7/26/2019 Thesis 8-2-16.docx

    62/129

    -3

    -2

    -1

    0

    1

    2

    3

    4

    5

    -3

    -2

    -1

    0

    1

    2

    3

    4

    5

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICE

    NON COMMERCIALNE TLONGPOSITION

    -2

    -1

    0

    1

    2

    3

    -2

    -1

    0

    1

    2

    3

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICE

    TRADE WEIGHTED DOLLAR INDEX

    -3

    -2

    -1

    0

    1

    2

    3

    -2

    -1

    0

    1

    2

    3

    4

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICE

    S & P 500

    -5

    -4

    -3

    -2

    -1

    0

    1

    2

    3

    -3

    -2

    -1

    0

    1

    2

    3

    4

    5

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI ASIS

    -4

    -3

    -2

    -1

    0

    1

    2

    3

    4

    -3

    -2

    -1

    0

    1

    2

    3

    4

    5

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICE

    OECD IMPORTS

    0

    20

    40

    60

    80

    100

    120

    140

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OIL SPOT PRICE

    -4

    -3

    -2

    -1

    0

    1

    2

    3

    4

    -3

    -2

    -1

    0

    1

    2

    3

    4

    5

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICE

    OECD CONSUMPTION

    -4

    -3

    -2

    -1

    0

    1

    2

    3

    -3

    -2

    -1

    0

    1

    2

    3

    4

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICES

    OECD IN!ENTOR"

    -2#0

    -1#5

    -1#0

    -0#5

    0#0

    0#5

    1#0

    1#5

    2#0

    -3

    -2

    -1

    0

    1

    2

    3

    4

    5

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICE

    OPEC PRODUCTION

    -4

    -3

    -2

    -1

    0

    1

    2

    3

    -3

    -2

    -1

    0

    1

    2

    3

    4

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICE

    TOTALOIL RIGCOUNT

    -1#2

    -0#8

    -0#4

    0#0

    0#4

    0#8

    1#2

    1#6

    2#0

    2#4

    2#8

    -4

    -3

    -2

    -1

    0

    1

    2

    3

    4

    5

    6

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTICRUDE OILSPOTPRICE

    INDUSTRIALPRODUCTION IN INDIA

    -1#5

    -1#0

    -0#5

    0#0

    0#5

    1#0

    1#5

    2#0

    -3

    -2

    -1

    0

    1

    2

    3

    4

    96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

    WTI CRUDE OILSPOTPRICE

    INDUSTRIALPRODUCTION IN CHINA

    62

  • 7/26/2019 Thesis 8-2-16.docx

    63/129

    T!)7$ 2. D$"'ri:#i$ "#!#i"#i'" %& &9n;!$n#!7, &in!n'i!7 !n; ":$'97!#i$ !ri!)7$"

    D$#$rin!n#" M$!n S#;. D$. S=$? >!>H*+> ??!>+A ??!+*@+ *!*+?+>

    :$8C $R:%"CTI:5 *!**?A+ *!*?>>*> *!>A?? >!@@A?> @!*+A@+ ?>!+@+ ?>!@H *!*?>

    :8C% I5085T:RE *!**** *!**@H *!>*?H !>H ++!A++> +!>>A? ?! *!*?>>

    :8C% C:5&"M$TI:5 *!***?A *!*A+ *!A?+A A!@@+A ?!H>>>@ A!@?*@ +A!>+AA@ *!A>A*+

    :8C% 58T IM$:RT& *!***>? *!*+@ *!*A@>* A!>+ A!*A*+H ?H!HH +*!?*@ *!>

    I5%"&TRI34

    $R:%"CTI:5 I5%I3 *!**@>A *!*>@+ *!H*A? !@+*H *!?A -2.012 A!@ *!A+>?H

    I5%"&TRI34

    $R:%"CTI:5 C6I53 *!**A@ *!*>+? *!*A>> +!H>@*+ ??!>>A +!*?+@?> !?HH *!?+HH?

    &=$ >** *!**>H *!*>*+ *!HH>?> !>+*?@H >?!H+? ?+!>A ?+!H+*AA *!?H?

    %:443R I5%8F *!***+> *!*?AA *!+**@ !**++ +!?+? ?*!+*A> ?*!>*@ *!+*H?A

    63

  • 7/26/2019 Thesis 8-2-16.docx

    64/129

    ?A M B3&I& *!*?*>@A *!@@?> *!*@H H!@@H ?!@>>+ ?!*H A*!>A *!??*?@

    5C 58T 4:5# $:&ITI:5 *!**@AA *!>*@H *!@ ?*!*@H >>@!@@ ?!AA++ ?!?>+ *!??+H

    5otes' ?* significance, > significance, ? significance

    64

  • 7/26/2019 Thesis 8-2-16.docx

    65/129

    2.6.1 H*:%#+$"i" 1 T$"# &%r N%n-7in$!ri#*

    The nonlinearity in the returns series is tested using B%& test! The results of the B%& test are

    presented in the Table A!! The test statistics are presented in the same table! The test statistics

    are significantly greater than the critical alues! Thus, the null hypothesis of I!%!% is re1ected!

    The residuals of linear :4& model are also tested for nonlinearity and found that the residual

    series is nonlinearly dependent on the oil prices!

    T!)7$ 2. BDS T$"# R$"97#"

    %imension B%& &tatistic $rob!

    A *!*?@@@ *!**+

    + *!*A?H?? *!**@

    *!*A> *!**@

    > *!*A>?>? *!*??

    *!*AA?H *!*A?

    5ote' This table reports the results of B%& test!

    %enotes the re1ections of the null hypothesis at ?* significance leelP %enotes the re1ections of the null

    hypothesis at > significance leelP %enotes the re1ections of the null hypothesis at ? significance leel!

    The study considered two states in the Markoregime switching model! These two states

    represent the high and lowolatility regimes in the model! Chen and Chen (A**H) used two

    state Markoregime switching model to differentiate bull and bear markets! 4i (A**H) used two

    state MarkoRegime &witching Model to distinguish low and high uncertainty in stock markets!

    In our model the two regimes are classified as high and lowolatile regimes! The regimes are

    classified based on their model ariance! 6igher the model ariance the state termed as high

    olatility state and lower the model ariance the state termed to be low olatility state!

    2.6.2 H*:%#+$"i" 2 E&&$'# %& S:$'97!#i%n %n #+$ Oi7 Pri'$ M$!"9rin8 #+$ M!r8in!7 E&&$'#

    %& S:$'97!#i%n

    65

  • 7/26/2019 Thesis 8-2-16.docx

    66/129

    The impact of speculatie factors on the 9TI oil price is estimated using Markoregime

    switching regression model! &tate ? is lowolatile regime and state A is higholatile regime!

    The results are presented in Table A!>! It is obsered from the results that in lowolatile regime

    the impact of noncommercial trader2s n$# 7%n8 :%"i#i%non the 9TI oil price is insignificant,

    whereas in higholatile regime it is significant at ?*! This indicates that speculatie factor has

    a negligible effect on the oil price! The difference in the model ariance between the two states is

    minimal which indicates that the ariations in the crude oil price eplained by speculatie factor

    in both the states is nearly the same and hae minimal significance!

    The duration of the regime ? is ?H>!H and regime A is >!A! This shows that olatility of crude oil

    conditioned on speculatie factor remained in the lowolatile regime for most of the times! The

    leel of speculation captures by noncommercial trader2s net long position is consistentirrespectie of the regime!

    T!)7$ 2.@ E"#i!#i%n r$"97#" %& M!r=%-R$8i$ S significance, ? significance

    Fi8 2.@ C%n;i#i%n!7 $!n", V%7!#i7i#i$" !n; S%%#+$; #r!n"i#i%n :r%)!)i7i#i$" &%r %;$7 1

    66

  • 7/26/2019 Thesis 8-2-16.docx

    67/129

    0 50 100 150 200 250-0.4

    -0.2

    0

    0.2

    0.4

    Explained Variable #1

    0 50 100 150 200 2500.0725

    0.073

    0.0735

    0.074

    0.0745

    Conditional Std of Equation #1

    0 50 100 150 200 2500

    0.5

    1

    1.5

    TimeSm

    oothed

    States

    Probabilities

    State 1

    State 2

    It is obsered from the

  • 7/26/2019 Thesis 8-2-16.docx

    68/129

    B!"i"has insignificant impact on the 9TI oil price in both higholatile regime and lowolatile

    regime! This indicates that basis hae no significant role in the determination of oil price during

    all times market phases (5ormal, Bearish, and Bullish)!

    The impact of SP @00on 9TI oil price is insignificant in lowolatile regime, whereas in high

    olatile regime it is significant at ?! This indicates that stock market inde eplains crude oil

    price only in the times of normal market phases! But, during bullish and bearish market periods

    stock market inde fails to eplain the dynamics in the oil price! It is epected to hae an inerse

    relationship with oil price, but a positie relationship is obsered between stock market inde

    and oil price! Rise in the price of oil might be due to the increased consumption of the crude oil

    during economic epansion!

    The duration of the regime ? is ?!A and regime A is >!*A! This shows that olatility of the crude

    oil conditioned on combined speculatie and financial factor is high, that is oil price is more

    fluctuating! The leel of speculation is significant irrespectie of the regime when controlled for

    financial factors! The difference in the model ariance between the two states indicates that the

    ariations in the crude oil price is better eplained by speculatie and financial factor in high

    olatile states compared to the lowolatile state!

    T!)7$ 2.6 E"#i!#i%n r$"97#" %& M!r=%-R$8i$ S

  • 7/26/2019 Thesis 8-2-16.docx

    69/129

    Model ariance *!**A+++ *!**@?

    8pected duration of Regime ?!A >!*A Time periods

    Transition $robabilities Matri (std! error, palue) * *!?A+>

    ? *!H>

    5otes' denotes ?* significance, denotes > significance, denotes ? significance

    Fi8 2.6 C%n;i#i%n!7 $!n", V%7!#i7i#i$" !n; S%%#+$; #r!n"i#i%n :r%)!)i7i#i$" &%r %;$7 2

    0 50 100 150 200 250-0.4

    -0.2

    0

    0.2

    0.4

    Explained Variable #1

    0 50 100 150 200 2500.04

    0.05

    0.06

    0.07

    0.08

    Conditional Std of Equation #1

    0 50 100 150 200 2500

    0.5

    1

    TimeSmoothedStatesProbabilities

    State 1

    State 2

    It is obsered from the that there is a freDuent switchoer between the two states! The

    regression results indicates that speculation combined with financial factors eplains the changes

    in the 9TI oil price

    69

  • 7/26/2019 Thesis 8-2-16.docx

    70/129

  • 7/26/2019 Thesis 8-2-16.docx

    71/129

    OPEC :r%;9'#i%nhas significant positie relationship in both high and lowolatile regimes! It

    implies that the :$8C production is lower than the eDuilibrium leel! :ur result suggests that

    :$8C production is one of factor in determining the oil price in all market phases! The impact

    OECD "#%'=on oil price in both high and lowolatile regimes is significant at ?! In low

    olatile period the relationship is negatie, whereas in higholatile period it is positie!

    OECD '%n"9:#i%nhas insignificant impact on the 9TI oil price in lowolatile regime,

    whereas in higholatile regime it is significant at ?! This indicates that :8C% consumption

    has significant role in the oil price determination only during bearish and bullish market phases!

    Its impact is insignificant during normal market phase!

    The impact of OECD n$# i:%r#"on 9TI oil price in lowolatile regime is insignificant,

    whereas in higholatile regime it is significant at ?! The relationship between the crude oil and

    :8C% import is negatie! :8C% consumption plays a significant role in determination of oil

    price only during bearish and bullish market phases!

    The relationship between the crude oil and In;9"#ri!7 :r%;9'#i%n %& In;i!is negatie and is

    significant at ? in higholatile regime! In lowolatile regime, it is positie and is significant

    at ?! This implies that Industrial production of India plays a key role in estimating the 9TI

    crude oil price!

    The impact of In;9"#ri!7 :r%;9'#i%n %& '+in!is insignificant in lowolatile regime, whereas it

    is negatie and is significant at ? in higholatile regime! This implies that Industrial

    production of China plays a significant role in determination of the oil price only during bearish

    and bullish market phases!

    The duration of the state ? is ?! and state A is ?!?! This indicates that the olatility of crude

    oil is moderately eplained by combined speculatie, fundamental, and financial factors! The

    leel of speculation is significant in higholatile regime indicating that speculatie factor hassignificance in determining the oil price only in higholatile phase when combined with other

    factors!

    71

  • 7/26/2019 Thesis 8-2-16.docx

    72/129

    The difference in the model ariance between the two states indicates that the ariations in the

    crude oil price is better eplained by speculatie, fundamental, and financial factors in low

    olatile states compared to the higholatile state!

    T!)7$ 2. E"#i!#i%n r$"97#" %& M!r=%-R$8i$ S

  • 7/26/2019 Thesis 8-2-16.docx

    73/129

    Model ariance *!***@ *!*****>

    8pected duration of Regime ?! ?!? Time periods

    Transition $robabilities Matri (std! error, palue) *!@ *!A

    *!* *!+

    5otes' ?* significance, > significance, ? significance

    It is obsered from the graphs that the transition between the states is less compared to the

    preious case!

    Fi8 2.C%n;i#i%n!7 $!n", V%7!#i7i#i$" !n; S%%#+$; #r!n"i#i%n :r%)!)i7i#i$" &%r %;$7

    73

  • 7/26/2019 Thesis 8-2-16.docx

    74/129

    0 50 100 150 200 250-0.5

    0

    0.5

    Explained Variable #1

    0 50 100 150 200 2500

    0.05

    0.1

    Conditional Std of Equation #1

    0 50 100 150 200 2500

    1

    2

    TimeSm

    oothed

    States

    Probabilities

    State 1

    State 2

    2. S9!r*

    This paper has analy-ed the impact of fundamental, financial and speculatie factors on the 9TI

    crude oil prices in high and low olatility periods using regimeswitching methodology!

    8mpirical results of 6ypothesis A suggest that speculation has a minimal effect on 9TI oil price!

    6ence, speculatie factors alone cannot eplain oil price changes! Minimal difference in ariance

    in states suggested that speculation does not change hugely across states! The estimates from

    6ypothesis + indicated that speculatie factors (5oncommercial traders net long position) when

    controlled for financial factors significantly eplain the 9TI oil price in both the regimes! This

    implies that speculation has greater eplaining power when incorporated with other factors!

    It is also obsered from the results (6ypothesis ) that the speculatie ariable, when controlled

    for fundamental and financial factors hae a significant role in eplaining the 9TI oil price in

    high olatility state! 3t low olatility regimes, fundamental, and financial hae significant role in

    price discoery of oil price! 9hile at the high olatility regimes, we obsere that factors

    pertaining to speculation combined with financial and fundamental factors hae a significant

    effect on the oil price!

    74

  • 7/26/2019 Thesis 8-2-16.docx

    75/129

    The results are of greater significance for the policy makers and regulators! The three important

    findings from the study are' (i) &peculation has minimal eplanatory powerP (ii) &peculation

    affects oil price positiely during lowolatile state and negatiely during higholatile state

    when controlled for financial ariablesP and (iii) &peculation has significant impact on oil price

    only in high olatility regime!

    The results suggest that the effect of speculation on oil price can only be seen in higholatile

    regimes! 9hereas, during lowolatile regimes, supply and financial factors plays a significant

    role in eplaining oil price! 6ence, it reDuired for the regulators and policy makers to look into

    supply dynamics, financial factors as well as speculatie factors of the oil prices based on the

    current in order to track or predict the moements of the oil price!

    75

  • 7/26/2019 Thesis 8-2-16.docx

    76/129

  • 7/26/2019 Thesis 8-2-16.docx

    77/129

    The research Duestions that are addressed in this chapter are as follows'

    ?! %oes comoement eists between the oil price and 8change rates of the oilimporting countries

    o 9hat is the strength of the comoement between the oil price and

    8change rates

    o 6ow it eoles across freDuency and oer time

    A! %oes comoement eists between the oil price and &tock Indices of the oil

    importing countries

    o 9hat is the strength of the comoement between the oil price and &tock

    indices

    o 6ow it eoles across freDuency and oer time

    This study focuses on the ma1or oilimporting countries, thus proiding holistic picture about

    how the oilimporting countries and their aried importing leelsDuantity impacts arious

    macroeconomic indicators! 3s oil being the one of the primary source of energy in these

    countries, the price of oil has a ma1or influence on the economy and that has arious effects on

    different macroeconomic ariables! Traders and institutional inestors hae aried inestment

    hori-ons, so both hae different risk profiles! 6ence, in the current study, the relationshipco

    moement between (i) the oil price and 8change rates and (ii) the oil price and &tock Indices is

    inestigated from traders and institutional inestor2s perspectie!

    "sing waelets, we can capture both shortrun and longrun moements in time and freDuency

    space! The study of the comoement between (i) the oil price and 8change rates and (ii) oil

    price and &tock Indices returns is crucial for risk assessment in financial industry! By using

    waelets, we study the lifecycle of comoements between ariables, which is essential for risk

    assessment to traders and institutional inestor2s! preious &tudies hae used waelet to measure

    77

  • 7/26/2019 Thesis 8-2-16.docx

    78/129

    the comoements across different ariables (Reboredo and Castro,A*?+a, bP #raham et al!,

    A*?+P Tiwari et al!, A*?+a, b)!Traditional time series models only look into the time scale of the

    ariables, whereas waelets consider both time and freDuency domain! There by using waelets

    eliminate the limitations of the standard time series model!

    Fi8 .1 M!>%r %i7 i:%r#in8 '%9n#ri$"

    &ource' 8I3 Top world oil net importers, A*?AP

    The

  • 7/26/2019 Thesis 8-2-16.docx

    79/129

  • 7/26/2019 Thesis 8-2-16.docx

    80/129

    T!)7$ .1 R$'$n# 7i#$r!#9r$ %n r$7!#i%n"+i: )$#G?7

    A**AG+

    Russia 03R and Co

    integration

    Russian economy is influenced significantly by

    fluctuations in the oil prices and the real echange rate

    %awson (A**H) ?@@?M?7

    A**>M?

    %ominican

    Republic

    Multiariate

    regression model

    and Cointegration

    Increasing oil prices depreciate the peso

    BenassyGuere

    et al! (A**H)

    ?@HM?7

    A**M??

    "&3 Cointegration and

    causality tests

    Causality runs from the oil to the dollar!

    Chen and Chen ?@HAM?7 #H Countries Cointegration Real oil prices may hae been the dominant source of

    80

  • 7/26/2019 Thesis 8-2-16.docx

    81/129

  • 7/26/2019 Thesis 8-2-16.docx

    82/129

  • 7/26/2019 Thesis 8-2-16.docx

    83/129

  • 7/26/2019 Thesis 8-2-16.docx

    84/129

  • 7/26/2019 Thesis 8-2-16.docx

    85/129

    0ene-uela significant negatie impact on the crude oil prices

    Reboredo et al!

    (A*?)

    /anuary A*** to

    May A*?A

    8"R:$835

    "5I:5(8"R),

    3ustralia

    (3"%), Canada

    (C3%), /apan

    (/$E), Meico

    (MF5),

    5orway (5:.)

    and the "nited

    .ingdom(#B$)

    %etrended cross

    correlation analysis

    The crosscorrelation analysis indicated that in longer

    time scales, oil price7echange rate correlations were

    negatie and low

    &econd, negatie dependence between oil and the "&

    dollar increased after the onset of the global financial

    crisis for all time scales

    Compilation of 4iterature' 3ppended to the preious work of Tiwari et al! (A*?+)!8isting literature summari-es that oil price increase has significant negatie effect on stock returns (/ones and .aul, ?@@P &adorsky, ?@@@P .ilianand $ark, A**HP 5andha and

  • 7/26/2019 Thesis 8-2-16.docx

    86/129

  • 7/26/2019 Thesis 8-2-16.docx

    87/129

  • 7/26/2019 Thesis 8-2-16.docx

    88/129

  • 7/26/2019 Thesis 8-2-16.docx

    89/129

  • 7/26/2019 Thesis 8-2-16.docx

    90/129

  • 7/26/2019 Thesis 8-2-16.docx

    91/129

    $ortugal, and the ".

    China, India, and

    Russia

    03R The impact of oil price shocks on stock prices in these l

    5I8s is mied, partly in contrast to the effects on the "&

    deeloped countriesL stock markets

    #il3lana and

    Eaya (A*?)

    /anuary A*** to

    %ecember A*??

    5igeria 4ong memory

    regression

    model

    The oil prices acting as a weakly eogenous ariable!

    &ignificant eidence of a positie relationship between t

    two ariables found though with a shortmemory effect

    Madaleno and

    $inho (A*?)

    %ecember

    ?@@A to ?>

    :ctober A*?A

    9orld indices (M&CI

    By &ectors)

    Continuous time

    waelets

    6igher coherence among series for higher scales

    supporting the interdependence hypothesis, showing

    long run market dynamics are more uncertain! Bidirec

    relationship between both series for large time hori-ons

    :lson et al!

    (A*?)

    ?st/anuary

    ?@>to Ath3pril

    A*?+

    "& Multiariate

    olatility

    #3RC6 models

    such as B8..,

    diagonal,

    constant

    conditional

    4ow &=$ >** returns cause substantial increases in the

    olatility of the energy inde! 9eak response from &=$

    olatility to energy price shocks

    91

  • 7/26/2019 Thesis 8-2-16.docx

    92/129

    correlation, and

    dynamic

    conditional

    correlation

    &ukcharoen et al!

    (A*?)

    H /anuary ?@A

    to +?%ecember

    A**H

    Canada, )

    ?>@'?*7

    A*?+'?A

    "&3 Timeseries The oil price is persistent and endogenous predictor ari

    and negatie oil prices predict the "& stock returns mor

    92

  • 7/26/2019 Thesis 8-2-16.docx

    93/129

  • 7/26/2019 Thesis 8-2-16.docx

    94/129

  • 7/26/2019 Thesis 8-2-16.docx

    95/129

  • 7/26/2019 Thesis 8-2-16.docx

    96/129

  • 7/26/2019 Thesis 8-2-16.docx

    97/129

    6ypothesis of the study of the effect of the 9TI crude oil prices on macroeconomic indicators

    of the oilimporting countries is formulated as follows'

    The comoement between the crude oil prices on 8change rate of the oilimporting as follows'

    H00: t!e co!erence coeicients bet2een t!e crude oil "rices and E+c!an%e rates are 3ero

    The comoement between the crude oil prices on &tock Inde of the oilimporting countries as

    follows'

    H01: t!e co!erence coeicients bet2een t!e crude oil "rices and Stoc* indices are 3ero

    .@. M$#+%;%7%8*

    9e used continuous waelet transform in this study to eamine the macroeconomic series for

    orthogonal waelet bases, where the choice is in between discrete and continuous! In analy-ing

    macroeconomic data the waelet transform proides a three dimensional diagram that illustrates

    time series information at different freDuencies, time, and strength! 9here the freDuency could

    range from low to high, time could range from of short to long term and finally the strength of

    association measured by color coding! In this study, we follow #rinsted et al! (A**) framework

    of continuous waelet transform (C9T), cross waelet transform (F9T), and waeletcoherency (9TC)!

    [email protected]. C%n#in9%9" !$7$# Tr!n"&%r

    3 waelet is a function with -ero mean and that is locali-ed in both freDuency and time! The

    C9T of a time series ( , nW ?,J, 5) with uniform time steps , is defined as the

    conolution of with the scaled and normali-ed waelet, defined as

    97

  • 7/26/2019 Thesis 8-2-16.docx

    98/129

    ( )( ) * !

    (n

    + n

    n

    t n n t 2 s +

    s s

    =

    (+!?)

    9e define the waelet power as

    A

    ( )n+

    2 s!The comple argument of

    ( )n+

    2 scan be interpreted as

    the local phase! &ince the waelets are not completely locali-ed in time, C9T has edge effects to

    addresses Cone of Influence (C:I) is introduced, where edge effects are ignored! The statistical

    significance of waelet power can be assessed relatie to the null hypotheses that the signal is

    generated by a stationary process with a gien background power spectrum ( )! 9e follow

    the procedure used by Torrence and Compo (?@@) for data generation using MonteCarlo

    simulation process! 3ccording to Torrence and Compo (?@@), regarding the white noise and red

    noise waelet power spectra, under the null, the corresponding distribution for the local waelet

    power spectrum at each time n and scalesis as

    (+!A)

    [email protected]. !$7$# '%+$r$n'$ (TC

    9aelet coherence (9TC) is used as a tool for finding the relationships between two series! The

    relationship of these series is found through freDuency bands and time interals! The waelet

    coherence can be seen as a locali-ed correlation coefficient in time freDuency space! 9aelet

    coherence of the change in the oil price and macroeconomic indicators differential series can be

    defined as'

    98

  • 7/26/2019 Thesis 8-2-16.docx

    99/129

  • 7/26/2019 Thesis 8-2-16.docx

    100/129

    T!)7$ . Fr$9$n'* i;$n#i&i'!#i%n

    Ti$ S'!7$" (In D!*" Fr$9$n'i$" Fr$. D$"'ri:#i%n"

    A 7

    6igh &hort Run ?

    ? +A

    Medium Medium Run+A 7

    7 ?A

    ?A 7 A>

    4ow 4ong RunA> ?*A

    T!)7$ . !$7$# In&$r$n'$

    :il price

    =

    8changerates &tock

    Indices3rrowspointingtowards

    Right In t!e "!ase7

    Cyclical effect 7 $ositie 8ffect

    "p 8change rates &tockIndices

    4agging :il price

    %own 8change rates &tockIndices

    4eading :il price

    4eft Anti4 out o t!e "!ase7 "p 8change rates &tockIndices

    4eading :il price

    100

  • 7/26/2019 Thesis 8-2-16.docx

    101/129

    3nti cyclical effect 7 5egatie8ffect

    %own 8change rates &tockIndices

    4agging :il price

    The time scales are classified as 6igh, Medium, and 4ow freDuency time scales! 9here 6igh

    represents the shortrun dynamics, Medium represents the Medium run and the 4ow represents

    4ong run! 3rrows pointing to the right indicates that the ariables, i!e!, benchmark oil prices andechange