FUTURES TRADING IN SELECTED NON-AGRICULTURAL COMMODITIES: A POST 2008 STUDY

Transcription

1 FUTURES TRADING IN SELECTED NON-AGRICULTURAL COMMODITIES: A POST 2008 STUDY A THESIS REPORT Submied by KALEEL NISHA Under he guidance of Dr. HAIDER YASMEEN in parial fulfillmen for he award of he degree of MASTER OF PHILOSOPHY in MANAGEMENT December 202

2 ii

3 iii

4 iv ONAFIDE CERTIFICATE Cerified ha his hesis repor FUTURES TRADING IN SELECTED NONAGRICULTURAL COMMODITIES: A POST 2008 STUDY is he bonafide work of Ms. KALEEL NISHA (RRN ) who carried ou he hesis work under my supervision. Cerified furher, ha o he bes of my knowledge he work repored herein does no form par of any oher hesis repor or disseraion on he basis of which a degree or award was conferred on an earlier occasion on his or any oher candidae. SIGNATURE Dr. HAIDER YASMEEN SUPERVISOR Professor Crescen usiness School.S. Abdur Rahman Universiy Vandalur, Chennai SIGNATURE Dr. MIRZA S. SAIYADAIN PROFESSOR AND DEAN Crescen usiness School.S. Abdur Rahman Universiy Vandalur, Chennai

5 v ASTRACT Derivaives are financial insrumens which do no have a value of heir own, bu derive heir value from he underlying asse like currency, commodiy or sock. The basic purpose of derivaives is hedging he risk of flucuaions in price of he underlying commodiy by aking posiion. u, in oday s world, hese derivaives have become an invesmen avenue for many invesors. Due o his, derivaive rading has an impac on he spo marke. Hence, sudy on derivaives has become an area of ineres o many researchers. This sudy makes an aemp o address some of he issues in Indian commodiy derivaive marke. In paricular, i analyzes he impac of fuures rading aciviy of gold, silver, copper and crude oil on heir respecive spo prices. Moreover, i deermines he relaionship beween fuures prices of hese commodiies. For his purpose, daa on daily spo prices, fuures rading volume and fuures prices for he seleced commodiies were obained. The year-wise volailiy in spo prices was ascerained using sandard deviaion of spo prices of hese commodiies. Uni roo es was carried ou o es he daa for saionariy. Muliple regression was used o find he impac of fuures rading volume of a given commodiy on is respecive spo prices. In addiion, i was also used o find ou which fuures conrac had higher impac on spo prices. The findings show ha he volume of near-monh fuures conrac had significan impac on he respecive spo prices of gold, silver and crude oil. u, volume of fuures conrac of copper did no have any significan impac on is spo prices. As he daa on fuures prices of he seleced commodiies were nonsaionary, Engle-Granger coinegraion es was used o assess he relaionship beween fuures prices of hese commodiies. On conducing pair-wise coinegraion es, i was found ha bi-direcional coinegraion exiss among all he pairs of seleced commodiies excep closing prices of gold fuures and silver fuures and gold fuures and copper fuures.

6 vi ACKNOWLEDGEMENT I would like o hank he managemen, he Vice Chancellor, he Regisrar, Dean (Academic Research) and our Dean (CS) for giving me an opporuniy o pursue M. Phil programme in his eseemed Universiy. I offer my profound graiude o my research supervisor, Dr. Haider Yasmeen for her valuable research guidance and advice hroughou my sudy. I am graeful o her for being a consan source of inspiraion and suppor for me. I would like o offer my sincere graiude o Dr. S. K. G. Ganesh who has given valuable advice and suggesions for my sudy. I would like o sincerely hank Dr. A. Hidhayahullah for his ideas and suggesions during he course of my sudy. I am also hankful o my faculy colleagues who have always been a source of moivaion and suppor for me.

7 vii TALE OF CONTENTS CHAPTER NO.. TITLE PAGE NO. ASTRACT v ACKNOWLEDGEMENT vi LIST OF TALES ix INTRODUCTION. OVERVIEW OF DERIVATIVES.2 HISTORY OF DERIVATIVES 3.3 INDIAN 4 COMMODITY DERIVATIVE MARKET.4 OJECTIVE Primary Objecive Secondary Objecives 7.5 SIGNIFICANCE OF THE STUDY 7.6 SCOPE OF THE STUDY 9.7 LIMITATIONS OF THE STUDY 9 REVIEW OF LITERATURE 2. REVIEWS ON IMPACT OF FUTURES TRADING ON SPOT MARKET 2.2 REVIEWS AMONG ON FUTURES CORRELATIONS PRICES 3 OF COMMODITIES RESEARCH GAPS 6 RESEARCH METHODOLOGY 7 3. RESEARCH DESIGN DATA USED FOR STUDY DATA ANALYSIS TECHNIQUE 7

8 viii 4. DATA ANALYSIS AND INTERPRETATION 9 4. VOLATILITY OF SPOT PRICES UNIT ROOT TEST MULTIPLE REGRESSION FOR SPOT 23 PRICES OF CRUDE OIL 4.4 MULTIPLE REGRESSION FOR SPOT 25 PRICES OF GOLD 4.5 MULTIPLE REGRESSION FOR SPOT 26 PRICES OF SILVER 4.6 MULTIPLE REGRESSION FOR SPOT 28 PRICES OF COPPER 4.7 RELATIONSHIP ETWEEN FUTURES PRICES OF CRUDE OIL, 30 GOLD, SILVER AND COPPER 4.7. Engle-Granger Tes Tes of coinegraion for fuures 32 prices of Crude Oil and Gold TREND ANALYSIS 33 CONCLUSION AND RECOMMENDATION SUMMARY OF RESEARCH RESULTS FINANCIAL IMPLICATIONS DIRECTIONS FOR FUTURE 37 RESEARCH 5.4 CONCLUDING REMARKS 38 REFERENCES 39 APPENDIX 40 LIST OF PULICATIONS 5 TECHNICAL IOGRAPHY 52

9 ix LIST OF TALES TALE NO. TALE NAME PAGE NO.. Value raded for seleced commodiies 8 4. Volailiy of spo prices Uni Roo Tes for Spo Prices of Crude Oil Resuls of Uni roo es for Crude Oil Prices and Volume Resuls of Uni roo es for Gold Prices and Volume Resuls of Uni roo es for Silver Prices and Volume Resuls of Uni roo es for Copper Prices and Volume Muliple regression for spo prices of Crude Oil Muliple regression for spo prices of Gold Muliple regression for spo prices of Silver Muliple regression for spo prices of Copper Resuls of Engle-Granger coinegraion es Resuls of Trend Analysis 34

10 . INTRODUCTION Derivaives have become an imporan invesmen avenue for many financial invesors. Wih he adven of elecronic rading in commodiy derivaives, invesors have also included commodiies in heir porfolio. This is eviden from he increase in rading volume in he commodiy derivaive marke and cash selemen dominaing he marke wih negligible physical selemen. Hence, an aemp is made o analyze wheher he fuures rading volume has significan impac on he spo prices of he underlying commodiy. In his chaper, an overview of derivaives, is evoluion, and developmens in Indian commodiy marke are discussed. The objecives of he sudy, significance, scope and limiaions are also included in his chaper.. OVERVIEW OF DERIVATIVES Derivaive is a financial insrumen whose value depends on price of anoher financial insrumen, ineres rae or indices. I is a form of alernaive invesmen. I has no value of is own and i is no a sandalone asse (John (2006)). In oher words, derivaives are producs whose values are derived from one or more basic variables called bases. These bases can be underlying asses (for example forex, equiy, ec), bases or reference raes. They are basically used o hedge risk. Hedging refers o a mehod of reducing he risk of loss caused by price flucuaion. The pracice of aking a posiion (buy or sell) in one marke o offse and balance he risk adoped by assuming a posiion (sell or buy) in an opposie marke is called hedging. In oher words, when a person wans o buy an asse in one marke on a fuure dae, he can look for anoher person who wans o sell he same asse in same marke on he same dae for an agreed-upon price and ener ino a conrac. Such a conrac is called forward conrac which is a ype of derivaive insrumen. This will offse he buyer s exposure o price flucuaions. The same can be done for sellers also. This is how derivaives can be used for hedging.

11 2 There are differen ypes of derivaive insrumens namely fuures, forwards, opions and swaps. Le us ake he case of fuures. In a fuures conrac, which is a derivaive insrumen here are wo paries o he conrac who commi o he ransacion of he underlying commodiy a a predeermined dae and price. In oher words, one pary agrees o buy and he oher agrees o sell he underlying commodiy on he specified fuure dae and price for a given quaniy and qualiy. Such fuures conracs are raded on sandardized fuures exchanges. Hence, fuures is a derivaive securiy which can be used o hedge risk. For example, suppose ha A wans o buy 5000 shares of a company six monhs from now and owns 5000 shares of same company worh Rs. 0,00,000 ha he wans o sell afer six monhs. Suppose ha hey boh agree on a sale price of Rs. 0,50,000. Now A and are said o have enered ino a fuures conrac, where A has a long posiion and has a shor posiion. Even if, afer six monhs he marke value of 5000 shares becomes Rs. 0,40,000, A is obliged o buy hose shares only a Rs. 0,50,000 and gains Rs 0,000. Thus has used derivaive o hedge he risk of flucuaion in share price. Opions and swaps are oher ypes of derivaive insrumens used for hedging. Opion is an insrumen ha conveys he righ bu no he obligaion o engage in a fuures conrac. A seller or a buyer of opion need no exercise he opion if i is no profiable for him/her whereas in fuures he conrac gives he holder he obligaion o abide by he erms of conrac. A swap is a derivaive in which counerparies exchange cerain benefis of one pary's financial insrumen for hose of he oher pary's financial insrumen. In swap here are differen ypes namely, ineres rae swap, currency swap, commodiy swap, ec.,(john (2006)). For example, a person who has acquired loan wih fixed ineres rae can swap i wih anoher person who has loan wih floaing ineres rae, provided, such a swap is beneficial or required for boh he paries. Though hese are oher derivaives in exisence, hey have no gained imporance in commodiies marke. They are prevalen only in sock marke and currency derivaive marke.

12 3 As seen from he above discussion, derivaives perform he imporan funcion of hedging o miigae he risk of flucuaion in spo prices. Moreover, hey ac as an indicaor of spo prices and he use of derivaives is no only limied o hedging bu also speculaion and arbirage. Hence here are hedgers, speculaors and arbirageurs who acively paricipae in he derivaive marke. The nex secion gives a brief evoluion of derivaives..2 HISTORY OF DERIVATIVES Russo e al., (2002) made a comparison on he evoluion of derivaives in Europe and US. From his sudy, i has been found ha fuures conrac for rading of ulip bulbs exised in Neherlands in 7h cenury. Fuures rading on meals and agriculural commodiies were cenered a London. Local commodiy exchanges sared funcioning o faciliae agriculural producion and markeing. Today, Eurex, European Energy Exchange, and so on are some of he world s leading derivaive exchanges siuaed in Europe. In Unied Saes, fuures rading daes back o 850s. These conracs were based on agriculural commodiies, peroleum producs and meals. Exchanges were organized in Chicago as i was a major disribuion cenre. Chicago oard of Trade, a fuures exchange, organized Chicago oard Opions Exchange 35 years ago which included derivaives in equiy. Consequenly, derivaives exchange in Unied Saes offers a mix of agriculural commodiies, peroleum producs and equiy derivaives. In 979, New York Sock Exchange, in which only financial securiies were raded, expanded ino fuures rading. Today here are COE Fuures Exchange, Chicago Mercanile Exchange, New York Mercanile Exchange and so on which are funcioning as derivaive exchanges in Unied Saes. Sarkar (2006) conduced a sudy on developmen of derivaives in India. The origin of derivaives rading in India daes back o nineeenh cenury and he ombay Coon Trade Associaion became an organised marke for commodiy fuures rading in 875. India was one of he larges markes for

13 4 commodiy fuures rading in early wenieh cenury. In 952, he governmen banned derivaive rading. There was also unorganized forward rading in sock marke, called badla rading. As derivaive insrumens are highly useful for hedging risk no only for individual invesors bu also for corporae, governmen of India ook seps o make organized derivaive marke for sock and for commodiies. This coninued when he ban on fuures rading was lifed and naional elecronic exchanges were inroduced. Various commiees were formed o review he process. As a resul derivaives rading was inroduced in NSE in 2000 for sock and Forward Marke Commission was se up in 2003 o regulae he Commodiy derivaive marke. Over he ime derivaives rading was inroduced in oher commodiies like jue, oilseeds, whea and bullion. A more deailed discussion on Indian commodiy derivaive marke is given below..3 INDIAN COMMODITY DERIVATIVE MARKET Indian marke has many avenues open for invesmen. Included in hese are commodiy derivaives and financial derivaives of recen origin. Alhough rading in commodiies exised in India from long ago, he scope of invesmen in commodiy marke was limied earlier. Wih he inroducion of elecronic rading in commodiy exchanges in 2003, he scope of invesmen in commodiies has become wider by seing up of hree naional level muli-commodiy exchanges, namely, Naional Muli Commodiy Exchange, Muli Commodiy Exchange, and Naional Commodiies and Derivaives Exchange. Derivaives are raded for many reasons. The mos imporan use of derivaives is hedging, where a rader can use derivaives o hedge he risk of flucuaions in spo prices of underlying commodiy. u, in pracice, derivaives are also used by speculaors who make profi or loss from anicipaed price movemens. I is ofen difficul o disinguish beween hedging and speculaion in a rade. u boh are imporan for an acive marke. There is anoher class of paricipans in his marke, namely, arbirageurs who make profi from discrepancies in prices prevailing in differen places. Like oher counries, hese hree classes of raders are presen in India

14 5 also, hough physical selemen of hese ransacions is very limied; mosly cash selemen akes place. In India, apar from numerous regional exchanges, as on dae here are five naionalized commodiy exchanges - Muli Commodiy Exchange (MCX), Naional Commodiy and Derivaives Exchange (NCDEX), Naional Muli Commodiy Exchange (NMCE), Indian Commodiy Exchange (ICEX) and Ace Derivaives and Commodiy Exchange (ACE). Around 03 commodiies are raded on hese exchanges. These commodiies are raded under various commodiy heads like agriculural commodiies, bullion, meals, energy, oil and oil producs, weaher and ohers. Indian commodiy fuures volume has grown o Rs rillion in he financial year o March 20 compared o rillion in he year o March Average monhly raded value sands a around Rs. 6 rillion (Mukherjee (20)). The rading hours for he commodiy exchanges in India is from 0.00 am o.30 pm for boh non-agriculural commodiies and agriculural commodiies from Monday hrough Friday. The rade imings for Saurdays are from 0.00 am o 2.00 pm. The rading holidays include he lis of holidays in he calendar year. There is a special session for Monday o Saurday from 9.45 am o 9.59 am o cancel any pending orders. Among he commodiies raded, he commodiies seleced for sudy include Crude Oil, Gold, Silver and Copper. Crude Oil is mixure of hydrocarbons ha exiss in liquid phase. Crude Oil is usually measured in barrels, which equals lires and raded in los of 00 barrels. India and China are he major buyers of oil in Asia as idenified by OPEC counries. The fuures conrac for Crude Oil expires in all monhs of he year usually on 20h of every monh. Fuures conrac for one monh o 6 monhs can be raded on commodiy exchanges. Gold is a commodiy which is raded under he commodiy head precious meals or bullion. India is he larges consumer of Gold in 200 wih an annual

15 6 demand of 963 onnes. Gold fuures conrac have expiry daes on he 5 h of conrac expiry monh. Gold fuures conrac expires in he monhs of February, April, June, Augus, Ocober and December. Gold fuures conrac which can be raded on exchanges include hose conracs ha are mauring in he firs neares monh o hose mauring in he fourh neares monh. Gold is usually raded for differen lo sizes like Gold peal for gm, Gold mini for 00 gm and Gold mega for 000 gms denoed as Gold. Under he precious meals caegory, Silver is also included. India s Silver demand averages 2500 ons per year and he counry s producion was around onnes in 200. Similar o Gold conrac, Silver conrac also expires on 5h of conrac expiry monhs. The conrac expiry monhs for Silver include, March, May, July, Sepember and December. Silver is raded in los of kg for Silver micro, 5 kgs for Silver mini and 30 kg for Silver mega denoed as Silver. Copper is aken under he meals caegory, for he presen sudy. February, April, June, Augus and November are he conrac expiry monhs for Copper. Copper conracs expire on he las dae of conrac expiry monh. I is raded in los of 250 kg in case of Copper mini and on in case of Copper mega denoed as Copper. Alhough here are many commodiies being raded on hese exchanges, he Forward Marke Commission acs as a regulaor and akes he responsibiliy of including or removing a commodiy from fuures rading by idenifying volailiy in spo prices due o fuures rading aciviy. This has even led o banning of fuures rading in cerain commodiies like urad, ur and whea. Wih all hese developmens peraining o commodiy derivaives, i has become an area of ineres o many researchers. Hence his sudy is also an aemp made o address some of he issues in Indian commodiy derivaive marke.

16 7.4 OJECTIVE.4. Primary Objecive The main objecive of he sudy is o analyze he impac of fuures rading volume of Gold, Silver, Copper and Crude Oil on heir respecive spo prices and idenifying he relaionship among heir fuures prices afer Ocober Secondary Objecives ased on he primary objecive, he following secondary objecives which are specific o he sudy are lised. To assess he year-wise volailiy in spo prices of Crude Oil, Gold, Silver and Copper. To es he ime series daa for saionariy. To sudy he impac of fuures rading volume of Gold, Silver, Copper, and Crude Oil on heir spo prices. To sudy he relaionship beween fuures prices of Gold, Silver, Copper and Crude Oil. To deermine he rend in spo prices and fuures prices of he seleced commodiies..5 SIGNIFICANCE OF THE STUDY The sudy on commodiy derivaives is essenial in India as commodiies conribued o 45% of GDP in 20 (Mukherjee (20)). There have been a significan number of sudies on commodiy derivaives in India bu many of hem have focused on agriculural commodiies and volailiy in heir prices due o fuures rading. Agriculural commodiies dominaed he Indian commodiy derivaive marke upo he year I is afer his period ha rading in non-agriculural commodiies like meals and energy producs sared gaining momenum. In spie of his fac, he sudy on commodiy derivaives of precious meals, meals

17 8 and energy producs raded on Indian commodiy exchanges are very limied alhough hese were analyzed by many researchers in foreign counries using fuures daa from heir respecive commodiy exchanges. Derivaive rading by invesors has led o price flucuaions of commodiies like urad, ur, as a resul, hese were banned for commodiy rading. The curren sudy hrows ligh on he impac of fuures rading on spo prices of gold, silver, copper and crude oil. Similarly, he findings of his sudy would also help in making policy decisions. Moreover, he seleced commodiies conribue o maximum value raded among oher commodiies as eviden from he following able. Table.: Value raded for seleced commodiies Commodiy Jan - 2 Value raded in Crore Rs. Feb - 2 Mar - 2 Apr -2 May - 2 Jun - 2 Jul - 2 Aug - 2 Gold Silver Meals Energy (23%) (26%) (8%) (7%) (22%) (29%) (2%) (4%) (2%) (26%) (9%) (5%) (9%) (27%) (23%) (6%) (24%) (26%) (9%) (9%) (25%) (26%) (7%) (2%) (22%) (8%) (6%) (25%) (9%) (22%) (7%) (27%) Toal Source: ased on daa for value-raded of commodiies obained from

18 9 Values wihin parenheses indicae percenage of value raded of each commodiy o oal value raded. The sudy on relaionship among fuures prices helps in undersanding he cross-marke rading dynamics. The sudy would faciliae fund managers in making beer invesmen decisions and framing good porfolio. The presen sudy highlighs he significance of fuures rading volume on heir spo prices. Hence his sudy can be used by invesors, commodiy rading companies and Governmen o undersand he impac of fuures rading volume on heir spo prices and also esimae spo prices..6 SCOPE OF THE STUDY The presen sudy uses daa on spo prices, fuures rading volume and fuures prices of he seleced commodiies from Muli Commodiy Exchange (MCX). This is due o he fac ha MCX ranks among he op 5 commodiy exchanges in he world for fuures rading in Gold and Crude Oil and s in Silver and Copper in India. The period of sudy ranges from Ocober 2008 o Augus 202 as here were many missing values prior o his period..7 LIMITATIONS OF THE STUDY The presen sudy has he following limiaions: Daa for he prices and volumes of seleced commodiies were aken only for five weekdays Monday o Friday o remove ouliers. Daa for Saurdays were no included because fuures rading aciviy was very limied as he working hours on Saurday is from 0.00 am o 2.00 pm. As he websie, was under consrucion, he individual value raded for Copper and Crude Oil under he meals and energy commodiies caegory were no able o be rerieved. Hence, value raded for meals and energy was given in Table..

19 0 Daa on fuures rading volume and prices for he longes conrac were no included in he sudy as mos of hese daa were missing. The sudy focused on only four non-agriculural commodiies. In spie of hese limiaions, he research carried ou is an earnes aemp o address he issues in Indian commodiy derivaive marke based on he research gaps idenified. Chaper 2 reviews he available lieraure and idenifies he research gaps.

20 2. REVIEW OF LITERATURE Many researchers have sudied he dynamics of commodiy derivaives marke under differen perspecives. This chaper reviews he available lieraure in his area and discusses heir findings. 2. REVIEWS ON IMPACT OF FUTURES TRADING ON SPOT MARKET Slade and Thille (2002) inegraed he produc marke and financial marke aspec of commodiy rading. They considered how produc-marke srucure and fuures-marke rading joinly affecs spo price levels and spo price volailiy. They have daa from London Meal Exchange and Raw Maerials Group. The period considered for sudy ranges from January 990 o January 999. The monhly-average of cash selemen price was used as spo price. Monhly average of daily sales of fuures conracs (in los, which is he conrac uni) divided by yearly Wesern world producion of he commodiy (also in los) and open ineres were used as a measure of rading aciviy. They have found a posiive relaionship beween produc-marke concenraion and price-levels using OLS mehod. They have also idenified a negaive relaionship beween rading aciviy and price level. A deailed sudy was conduced by Lokare (2007) who has made a criical evaluaion of commodiy derivaives marke in India. In his paper, he seeks o sudy how vibran is he fuures marke based on deph/breadh and liquidiy, volailiy of he fuures marke and effeciveness of he fuures marke in erms of price risk managemen using daa on spo prices and fuures prices. ased on his findings he has also suggesed he required policy response in he fuure. The liquidiy in Indian derivaive marke was analyzed based on he proporion of oal volume of rading of a commodiy o is oal producion in he counry. To assess he efficiency of derivaive marke, es of co-inegraion were used o check he co-movemen of spo and fuures prices. A marke is said o be efficien if he fuures prices are an unbiased esimaor of spo prices. The volailiy in fuures marke was deermined based on he raio of monh-wise

21 2 sandard deviaions of fuures prices and spo prices. The raio closer o one indicaes efficien marke, greaer han one indicaes speculaive aciviies and less han one indicaes less efficien marke. To sudy he effeciveness of fuures marke in price risk managemen, one can compare he sandard deviaion of basis risk wih ha of he spo prices in he mauriy monh. asis risk, also called marginal convenience yield is a measure of marginal sorage value of sorage of a commodiy calculaed by he spread beween he spo price and fuure price. Daa on spo prices and fuure prices was colleced for various commodiies from 999 o Liquidiy was found high in casorseed and soyabean compared o oher commodiies. Afer performing coinegraion ess, co-inegraing relaionships were winessed in commodiies like rice, whea, sugar, sesame seed and coon among primary commodiies; Gold, Copper, lead and in among meals; and bren Crude Oil among oils. Volailiy appeared o be low for gur, poao and rice. On sudying he volailiy, varying volailiy was observed for pepper, casorseed, rubber, sugar, coon, whea and musard. In case of meals, speculaive rading was displayed in Gold and Silver conracs. Differen commodiies showed varying levels of basis risk. Among he meals, basis risk was high in case of Copper, lead and in and low in case of aluminium and zinc. The same for Gold was moderae and ha of Silver was relaively low signifying ha Silver is less riskier han Gold. ased on hese findings, he policy responses include, bringing radiional players o formal marke, srenghen he inpu delivery sysem and spo/physical marke, shoren he exising long supply chain, graning indusry saus o commodiy secor and so on. The relaionship beween rading volume of fuures conrac of commodiies and heir spo prices was sudied by Nah and Lingareddy (2008). They analyzed his relaionship wih respec o hree commodiies urad, gram and whea. These hree commodiies were banned for fuures rading by he Indian Governmen when heir spo prices increased abruply in Their sudy focused on impac of fuures rading on spo prices and heir volailiies and comparing hem during he pre-ban period and pos-ban period. Daa peraining o prices, fuures and volume for he seleced commodiies were

22 3 obained for a period from January 200 o Augus For his purpose, linear regression and Granger causaliy es were used. Linear regression was used o sudy he facors influencing spo prices. Granger causaliy es was used o undersand if here is a bi-direcional impac flowing from one price o he oher and undersand he relaion beween volume and prices of urad, gram and whea. Their sudy found ha fuure rading has led o increase in spo price of urad bu no gram and whea. Their resuls indicae ha he average price levels as well as volailiies were higher in he period where fuures rading in hese commodiies was allowed. There has been a sharp fall in volailiy in he posban period which shows he exisence of fuures rading impac on spo prices. 2.2 REVIEWS ON CORRELATIONS AMONG FUTURES PRICES OF COMMODITIES The dependencies in commodiies marke and correlaions among heir prices were analyzed by Sieczka and Holys (2009). They have made an aemp o find he correlaion among commodiies like meals, fuels, plans and animals using correlaion marix. For his purpose 35 fuures conracs were invesigaed and daa colleced from Chicago oard of Trade (COT), Chicago Mercanile Exchange (CME), Inerconinenal Exchange (ICE), Kansas Ciy oard of Trade (KCT), London Meal Exchange (LME), Minneapolis Grain Exchange (MGEX), New York oard of Trade (NYOT), New York Mercanile Exchange (NYMEX), and Winnipeg Commodiy Exchange (WCE) for he period o They have consruced a minimal spanning ree for his purpose which showed exisence of srong correlaion among he prices of hese commodiies. I was also found ha he mean correlaion among commodiy prices have significanly increased over he years due o increase in he number of invesors in commodiy marke. Possibiliy exiss in relaed commodiies winessing co-movemen of heir fuures prices. Neverheless, similar co-movemen of fuures prices can also be seen in unrelaed commodiies. One such relaionship was evidenced by Chng (2009) who sudied he correlaion among fuures prices of naural rubber,

23 4 Palladium and gasoline on he Tokyo Commodiy Exchange ha are exposed o auomobile indusry. Daa relaing o he sudy were obained for he period from 3rd July 2000 o 3s March He used VAR and EKK-GARCH esimaions for pair-wise comparison beween each of naural rubber, Palladium and gasoline. Moreover, quasi vecor-moving-average (VMA) esimaion was used o ascerain wheher common indusry exposure or commodiy marke facor like producion cycle was more relevan a explaining hese ineracions. Afer performing hese esimaions i was found here exis bi-direcional relaionships among he pair-wise fuures prices of naural rubber, gasoline and palladium. The VMA resuls showed ha i was no commodiy marke facor bu common indusry exposure which was driving he cross-marke rading dynamics in hese commodiy fuures. The linkages among fuures prices of differen commodiies can also be used for oher purposes. One such use is he deerminaion of fuures price beyond he mauriy dae. Corazar and Eerovic (200) developed a modified mulicommodiy model o assess he prices of Copper and Silver fuures based on oil fuures prices. Their aim was o assess he prices of fuures beyond he longes mauriy conrac ha is raded in he marke. The model used variables like spo prices, fuures prices, ineres raes, convenience yield, ec., and imevarying risk premiums and differen volailiy specificaions are included. The daily fuures conracs raded a LME, NYMEX or ICE were used o ge he fuures prices of Copper, Silver, bren and Wes Texas Inermediae (WTI). I was found ha he modified mulicommodiy model helps in exrapolaing he long-erm Silver and Copper prices based on he Wes Texas Inermediae prices which were no possible in oher models which again esablishes he influence of fuures price of one commodiy on oher commodiies. The impac of fuures rading in commodiy markes of differen pars of he world on he fuures prices of commodiies in Indian commodiy exchanges is anoher area o be analyzed. This ineresing sudy was conduced by Kumar and Pandey (20) who esed inernaional linkages among commodiy

24 5 markes. Indian commodiy exchanges were compared wih commodiy exchanges ouside India for nine commodiies. Soyabean and corn were used under agriculural commodiies, Aluminium, Copper and Zinc under meals, Gold and Silver for precious meals and Crude Oil and Naural gas for energy producs. ased on he volume raded, daa for Indian commodiies were obained from Muli Commodiy Exchange and Naional Commodiy Derivaives Exchange and compared wih daa on commodiies obained from Chicago oard of Trade, London Meal Exchange and New York Meal Exchange. Mos of he daa were colleced in ime period beween 2004 and They have used Granger causaliy es, Johansen s coinegraion es, error correcion model and variance decomposiion echniques o invesigae he relaionship in fuures prices among hese markes. Volailiy spillover beween he Indian commodiy markes and is world couner-pars was esed using i-variae Generalized Auo Regressive Condiional Heeroskedaciy (GARCH) model. On carrying ou hese ess, hey have found ha he world markes have greaer impac on Indian marke. The resul of i-variae model indicaes ha here was bi-direcional effec beween fuures prices of commodiies from Muli Commodiy Exchange and London Meal Exchange. A very recen sudy was conduced by Ji and Fan (202) o show how price volailiy of Crude Oil marke is expanding o non-energy markes like meal marke, agriculure marke and aggregae non-energy marke. The volailiy spillover effecs beween Crude Oil marke and each of he oher commodiy markes and heir dynamic condiional correlaions were analyzed before and afer he 2008 economic crisis. They have used daa for he period ranging from 7 July 2006 o 30 June 200 from Wind daabase. Exponenial generalized auoregressive condiional heeroskedasiciy (EGARCH) model proposed by Nelson (99) and expanded o a mulivariae model was used. Their resuls indicaed ha price volailiy of Crude Oil marke had relaively higher influence on oher commodiy markes han ha of oher commodiy markes on Crude Oil marke. In addiion, his impac was comparaively greaer before he economic crisis. As far as condiional correlaion beween Crude Oil marke and oher

25 6 commodiy marke was concerned, his correlaion firs increased and hen decreased before he economic crisis bu here was a consan increase afer he economic crisis showing he influence of volailiy in Crude Oil fuures prices on volailiy of spo prices of oher commodiies. 2.3 RESEARCH GAPS From he papers reviewed so far, he following research gaps are idenified: The sudies conduced on non-agriculural commodiies raded on Indian commodiy exchanges are very limied. The relaionship beween fuures prices of commodiies raded on Indian commodiy exchanges is no explored. Mos of he sudies were concerned wih he impac of fuures rading aciviy on spo price volailiy and only few sudies deal wih spo price levels. Hence his sudy makes an aemp o address hese issues. The nex chaper gives an overview of he mehodology used in he sudy.

26 7 3. RESEARCH METHODOLOGY The mehodology adoped in he sudy is oulined in his chaper. I begins wih research design, followed by daa used for he sudy, and concludes wih echniques used o analyze he daa. 3. RESEARCH DESIGN This sudy used secondary daa for analyzing he impac of rading volume on spo prices and relaionship among fuures prices. Hence analyical research is used in his sudy. 3.2 DATA USED FOR STUDY Around 800 observaions were obained from rading days during he period Ocober 2008 o Augus 202. Daa was obained from MCX hav Copy available on is websie. The daa consised of daily spo prices, daily rading volume of fuures wih differen conrac expiry daes and daily closing prices of fuures conrac for all he seleced commodiies Crude Oil, Gold, Silver and Copper. Daa peraining o Saurdays were no included in he sudy as he rading hours were from 0.00 am o 2.00 pm, as compared o 0.00 am o.30 pm on Monday o Friday. 3.3 DATA ANALYSIS TECHNIQUE The year-wise volailiy in spo prices of Crude Oil, Gold, Silver and Copper were deermined using he sandard deviaion of heir spo prices. The saionariy of ime series daa was esed using Uni Roo Tes. If he daa was found o be saionary, Ordinary Leas Square Mehod can be used o es he impac of fuures rading volume on spo prices. If he daa was nonsaionary, firs-order differencing can be done o ransform he non-saionary daa o saionary daa. OLS mehod can be applied only when he daa is saionary.

27 8 In his sudy, muliple regression was used o es he impac of fuures rading volume on heir respecive spo prices of he seleced commodiies. To sudy he relaionship beween fuures prices of hese commodiies, EngleGranger coinegraion es was used. Uni roo es and Engle-Granger coinegraion es as illusraed in Gujarai and Sangeeha (2007) was used for analysis. Trend analysis was carried ou using linear regression model. Daa analysis was carried ou using SPSS 6.0 and Microsof Excel. Using he mehodology oulined in his chaper, daa on spo and fuures prices and fuures rading volume were analyzed. The following chaper deals wih he daa analysis and inerpreaion.

28 9 4. DATA ANALYSIS AND INTERPRETATION The impac of fuures rading volume on spo prices is analyzed used daa on spo prices of hese commodiies and heir fuures rading volume. To deermine he relaionship beween fuures prices of hese commodiies, daa on daily closing prices of neares fuures conrac are used. 4. VOLATILITY OF SPOT PRICES To assess he volailiy in spo prices of Crude Oil, Gold, Silver and Copper, heir year-wise sandard deviaion was calculaed and summarized below: Table 4.: Volailiy of spo prices Year Commodiy Crude Oil Gold Silver Copper (Upo Augus) From he above able, i can seen ha volailiy was highes for spo prices of Crude Oil in he year 2009 followed by 20. Spo prices of Gold show an increasing volailiy from he year 2009 o 20. Volailiy in spo prices of Silver was highes in he year 20. Among all he commodiies spo prices of Copper shows he leas volailiy. 4.2 UNIT ROOT TEST The daa on spo prices and rading volume of fuures of all he seleced commodiies were esed for saionariy using Uni Roo Tes.

29 20 Uni Roo Tes is explained as follows: A ime-series daa is said o be saionary if is auoregressive coefficien, ρ. If ρ =, i is he case of uni roo and becomes a random walk model wihou drif which is a non-saionary sochasic process. The value of ρ is obained by regressing he dependen variable on is lagged values of order. The equaion akes he form Y = ρy-+u - ρ (4.2.) Where uis he error erm. For heoreical reasons, Equaion 4.2. is manipulaed as follows: Y = δy-+u The null hypohesis is δ = 0, ha is here is a uni roo. If δ is negaive, Y is saionary. Taking he firs difference of spo prices of Crude Oil as he dependen variable ( Y) and is lagged values of lag lengh (Y-) as he independen variable, he following resuls were obained: Table 4.2: Uni Roo Tes for Spo Prices of Crude Oil (Consan) LAGS (Spo Price of Crude Oil,) Unsandardized Sandardized Sd. ea Error Dependen Variable: DIFF (Spo Price of Crude Oil,) From Table 4.2, i is eviden ha δ < 0, which infers ha he ime series daa on Spo prices of Crude Oil is saionary. u, i is no significan. Hence he value of ρ is sudied for significance and used for esablishing saionariy, he resuls of which are showed below. The oupus for Uni Roo Tes obained from SPSS are included in he Appendix.

30 2 Table 4.3: Resuls of Uni roo es for Crude Oil Prices and Volume S. No value p-value ** 0.000** Naure of daa Saionary Non-saionary ** Saionary ** Saionary ** Saionary ** Saionary ** Saionary ρ Variable Spo Price of Crude Oil Fuures price of Crude Oil Volume in los of one 3. monh fuures of Crude Oil Volume in los of wo 4. monh fuures of Crude Oil Volume in los of hree 5. monh fuures of Crude Oil Volume in los of four 6. monh fuures of Crude Oil Volume in los of five 7. monh fuures of Crude Oil **Significan a % level From Table 4.3, i can be seen ha daa on fuures prices of Crude Oil is non-saionary whereas oher daa are saionary. Table 4.4: Resuls of Uni roo es for Gold Prices and Volume S. No. Naure of daa Nonsaionary Nonsaionary ρ -value p-value. Spo Price of Gold ** 2. Fuures price of Gold ** ** Saionary ** Saionary ** Saionary Variable Volume in los of Gold 3. fuures mauring in firs neares monh Volume in los of Gold 4. fuures mauring in second neares monh Volume in los of Gold 5. fuures mauring in hird neares monh **Significan a % level Table 4.4 shows ha spo and fuures prices of gold are non-saionary whereas oher daa for Gold are saionary.

31 22 Table 4.5: Resuls of Uni roo es for Silver Prices and Volume S. Variable No.. Spo Price of Silver Fuures price of 2. Silver Volume in los of Silver fuures 3. mauring in firs neares monh Volume in los of Silver fuures 4. mauring in second neares monh Volume in los of Silver fuures 5. mauring in hird neares monh Naure of daa Saionary Nonsaionary ρ -value p-value ** ** ** Saionary ** Saionary ** Saionary **Significan a % level From he above able, i is eviden ha daa on fuures prices of Silver is non-saionary whereas oher daa are saionary. Table 4.6: Resuls of Uni roo es for Copper Prices and Volume S. No Naure of daa Saionary Nonsaionary Variable ρ -value p-value Spo Price of Copper Fuures price of Copper Volume in los of Copper fuures mauring in firs neares monh Volume in los of Copper fuures mauring in second neares monh Volume in los of Copper fuures mauring in hird neares monh ** ** ** Saionary ** Saionary ** Saionary **Significan a % level From Table 4.6, i can be seen ha he daa fuures prices of Copper are non-saionary whereas he oher daa are saionary.

32 MULTIPLE REGRESSION FOR SPOT PRICES OF CRUDE OIL The ime series daa for spo prices of Crude Oil and volume of fuures conrac are saionary. Muliple regression is used o sudy he relaionship beween spo prices of Crude Oil, considered as he dependen variable and heir respecive fuures volume wih differen mauriy, considered as independen variables. The muliple regression equaion akes he form, Y = α+βf + β2f2 + β3f3 +.+ βnfn +u Where Y Fi Spo price of Crude Oil in ime period Fuures raded volume of Crude Oil mauring in he ih neares monh in ime period α Consan β, β2,, βn Regression u Sochasic disurbance or error erm The following hypoheses were esed for finding he impac of fuures rading volume of Crude Oil on is spo prices and resuls obained are given in he able below. H : Fuures rading volume of Crude Oil has significan impac on is spo prices. H.: One monh fuures rading volume of Crude Oil has significan impac on is spo prices. H.2: Two monh fuures rading volume of Crude Oil has significan impac on is spo prices. H.3: Three monh fuures rading volume of Crude Oil has significan impac on is spo prices.

33 24 H.4: Four monh fuures rading volume of Crude Oil has significan impac on is spo prices. H.5: Five monh fuures rading volume of Crude Oil has significan impac on is spo prices. Table 4.7: Muliple regression for spo prices of Crude Oil (Consan) Volume in los of one monh fuures of Crude Oil Volume in los of wo monh fuures of Crude Oil Volume in los of hree monh fuures of Crude Oil Volume in los of four monh fuures of Crude Oil Volume in los of five monh fuures of Crude Oil R2 Adjused R2 N Unsandardized Sd. Error Sandardized ea Dependen Variable: Spo Price of Crude Oil From Table 4.7, i can be observed ha volume of one monh fuures conrac, volume of wo monhs fuures, volume of hree monhs fuures and volume of four monhs fuures conrac of Crude Oil are significan a % level. This implies ha volume of one monh fuures conrac, volume of wo monhs fuures, volume of hree monhs fuures and volume of four monhs fuures conrac of Crude Oil have a significan impac on spo prices of Crude Oil.

34 MULTIPLE REGRESSION FOR SPOT PRICES OF GOLD The ime series daa for spo prices of Gold is non-saionary whereas volume of Gold fuures conrac wih differen conrac expiry daes is saionary. Hence, firs-order differencing in spo prices of Gold is done o make i a saionary process. Muliple regression is used o sudy he relaionship beween spo prices of Gold and heir respecive fuures volume wih differen mauriy, considered as independen variables. The dependen variable is he Firs Differences of spo prices of Gold. The muliple regression equaion akes he form, Y = α+βf + β2f2 + β3f3 +.+ βnfn +u where Y Firs Difference of Spo price of Gold in ime period Fi Fuures raded volume of Gold mauring in he ih neares monh in ime period α Consan β, β2,..βn Regression u Sochasic disurbance or error erm The following hypoheses were esed for finding he impac of fuures rading volume of Gold on is spo prices and resuls obained are given in he able below. H 2: Fuures rading volume of Gold has significan impac on is spo prices. H 2.: Volume of Gold fuures mauring in he firs neares monh has significan impac on is spo prices. H 2.2: Volume of Gold fuures mauring in he second neares monh has significan impac on is spo prices.

35 26 H 2.3: Volume of Gold fuures mauring in he hird neares monh has significan impac on is spo prices. Table 4.8: Muliple regression for spo prices of Gold (Consan) Unsandardized Sandardized Sd. ea Error Volume in los of Gold fuures mauring in firs.00 neares monh Volume in los of Gold fuures mauring in second.004 neares monh Volume in los of Gold fuures mauring in hird.072 neares monh R2 Adjused R2 N Dependen Variable: DIFF (Spo Price of GOLD,) Table 4.8 shows ha volumes of Gold fuures mauring in he firs and second neares monh are significan, hereby having an impac on firs difference of spo prices of Gold. 4.5 MULTIPLE REGRESSION FOR SPOT PRICES OF SILVER The ime series daa for spo prices of Silver and volume of Silver fuures conrac are saionary. Muliple regression is used o sudy he relaionship beween spo prices of Silver, considered as he dependen variable and heir fuures volume, considered as independen variables.

36 27 The muliple regression equaion akes he form, Y = α+βf + β2f2 + β3f3 +.+ βnfn +u Where Y Fi Spo price of Silver in ime period Fuures raded volume of Silver mauring in he ih neares monh in ime period α Consan β, β2,..βn Regression u Sochasic disurbance or error erm The following hypoheses were esed for finding he impac of fuures rading volume of Silver on is spo prices and resuls obained are given in he able below. H 3: Fuures rading volume of Silver has significan impac on is spo prices. H 3.: Volume of Silver fuures mauring in he firs neares monh has significan impac on is spo prices. H 3.2: Volume of Silver fuures mauring in he second neares monh has significan impac on is spo prices. H 3.3: Volume of Silver fuures mauring in he hird neares monh has significan impac on is spo prices.

37 28 Table 4.9: Muliple regression for spo prices of Silver (Consan) Volume in los of Silver fuures mauring in firs neares monh Volume in los of Silver fuures mauring in second neares monh Volume in los of Silver fuures mauring in hird neares monh Unsandardized Sandardized Sd. ea Error R2 Adjused R2 N Dependen Variable: Spo Price of SILVER In he case of Silver, volume of fuures conrac mauring in he firs and second neares monh have significan impac on spo prices of Silver as shown in he able. 4.6 MULTIPLE REGRESSION FOR SPOT PRICES OF COPPER The ime series daa for spo prices of Copper and volume of Copper fuures conrac are saionary.

38 29 Muliple regression is used o sudy he relaionship beween spo prices of Copper, considered as he dependen variable and heir fuures volume, considered as independen variables. The muliple regression equaion akes he form, Y = α+βf + β2f2 + β3f3 +.+ βnfn +u Where Y Fi Spo price of Copper in ime period Fuures raded volume of Copper mauring in he ih neares monh in ime period α Consan β, β2,..βn Regression u Sochasic disurbance or error erm The following hypoheses were esed for finding he impac of fuures rading volume of Copper on is spo prices and resuls obained are given in he able below. H 4: Fuures rading volume of Copper has significan impac on is spo prices. H 4.: Volume of Copper fuures mauring in he firs neares monh has significan impac on is spo prices. H 4.2: Volume of Copper fuures mauring in he second neares monh has significan impac on is spo prices. H 4.3: Volume of Copper fuures mauring in he hird neares monh has significan impac on is spo prices.

39 30 Table 4.0: Muliple regression for spo prices of Copper (Consan) Unsandardized Sandardized Sd. ea Error Volume in los of Copper fuures.000 mauring in firs neares monh Volume in los of Copper fuures mauring in -2.78E-5 second neares monh Volume in los Copper fuures.008 mauring in hird neares monh R2 Adjused R2 N Dependen Variable: Spo Price of COPPER The analysis on spo prices of Copper and is fuures rading volume shows resuls differen from he resuls obained for oher commodiies. Volume of Copper fuures mauring in he firs neares monh does no have any influence on spo prices of Copper. The oher regression coefficiens are no significan a 5% level. Hence i can be concluded ha fuures rading volume of Copper does no influence spo prices of Copper. 4.7 RELATIONSHIP ETWEEN FUTURES PRICES OF CRUDE OIL, GOLD, SILVER AND COPPER The ime-series daa on fuures prices of Crude Oil, Gold, Silver and Copper were esed for saionariy. Daa on closing prices of fuures conrac wih neares mauriy were seleced for his analysis. Consisen wih many research sudies, hese daa were found o be non-saionary. Transforming he nonsaionary ime series daa o saionary daa using firs difference operaor may

40 3 yield confusing resuls. Hence, he relaionship beween he fuures prices of Crude Oil, Gold, Silver and Copper was esed pair-wise using Engle-Granger coinegraion es. The noaions used for fuures prices of he seleced commodiies are given below: FPCO - Closing price of one monh fuures conrac of Crude Oil FPG - Closing price of Gold fuures mauring in he firs neares monh FPS - Closing price of Silver fuures mauring in he firs neares monh FPC - Closing price of Copper fuures mauring in he firs neares monh 4.7. Engle-Granger Tes Engle-Granger or Augmened Engle-Granger es can be used o es he coinegraion beween pairs of fuures prices of he seleced commodiies. The condiion o be saisfied for using his es is ha, he ime series daa should be I(), ha is, i should be inegraed of order. Here, a regression equaion has o be esimaed by regressing he dependen variable on independen variable and heir residuals should be obained. Laer, hese residuals are esed for saionariy using Dickey-Fuller es. If he residuals are found o saionary, here exiss coinegraion beween he variables aken for sudy. Oherwise, here does no exis any coinegraion beween he wo variables. One variaion in he DF es is ha, Engle-Granger has obained he criical values of, o es for saionariy. Hence, he name Engle-Granger es. To es he exisence of coinegraion among he fuures prices of Crude Oil, Gold, Silver and Copper, he following hypoheses are used. H 5: There exiss coinegraion among he fuures prices of Crude Oil, Gold, Silver and Copper H 5.: There exiss coinegraion beween fuures prices of Crude Oil and Gold.

41 32 H 5.2: There exiss coinegraion beween fuures prices of Crude Oil and Silver. H 5.3: There exiss coinegraion beween fuures prices of Crude Oil and Copper. H 5.4: There exiss coinegraion beween fuures prices of Gold and Silver. H 5.5: There exiss coinegraion beween fuures prices of Gold and Copper. H 5.6: There exiss coinegraion beween fuures prices of Silver and Copper Tes of coinegraion for fuures prices of Crude Oil and Gold The closing prices of Crude Oil fuures (FPCO) are regressed on closing prices of Gold fuures (FPG) o ge he following regression equaion. The residuals obained are esed for saionariy and resuls are shown below. FPCO = FPG = (7.586) (47.375) On performing uni roo es for he residuals, he following resuls were obained: Δu = u- = (-4.05) R2 = 0.02 d =.992 (Durbin-Wason saisic) The Engle-Granger % criical τ (=) value is and 5% criical value is -.95 respecively. Since he compued τ (=) value is much negaive han his, he residuals obained from he regression of closing prices of Crude Oil fuures on closing prices of Gold fuures is I(0); ha is, hey are saionary. This shows ha here exiss coinegraion from closing prices of Crude Oil fuures o closing prices of Gold fuures. Hence here is a saic or long-run relaionship beween he wo.

42 33 Similarly, Engle-Granger coinegraion es was carried ou for he fuures prices of oher commodiies and heir resuls are summarized in he Table below. The oupus obained are included in he Appendix. Regressand Regressor Inercep Regression Coefficien -value τ-value for residuals Uni roo es for residuals Resul Table 4.: Resuls of Engle-Granger coinegraion es FPCO FPCO FPS FPC ** 52.4** ** -3.67** FPG FPS ** FPG FPC ** FPS FPC ** * Saionary Saionary Nonsaionary Nonsaionary Saionary Coinegraed Coinegraed No coinegraed No coinegraed Coinegraed *Significan a 5% level **Significan a % level From Table 4., i can concluded ha here exiss coinegraion beween closing prices of Crude Oil fuures and Gold fuures, Crude Oil fuures and Silver fuures, Crude Oil fuures and Copper fuures and Silver fuures and Copper fuures. u, here does no exis any coinegraion beween closing prices of Gold fuures and Silver fuures and Gold fuures and Copper fuures TREND ANALYSIS An aemp is made o find he rend in spo prices and fuures prices of crude oil, gold, silver and copper. Trend analysis is used o assess he rend in he ime series daa. I can be deermined by he regressing he ime series daa on ime. The regression equaion akes he form Y = α + β +u

43 34 Dependan variable in ime period Where Y Time period α Consan β Regression u Sochasic disurbance or error erm In he regression equaion, if he value of he regression coefficien is found o be posiive, he ime series daa shows an upward rend, oherwise, downward rend. The resuls of rend analysis are summarized as follows: Table 4.2: Resuls of Trend Analysis Dependan Variable Consan Regression coefficien - value Spo price of Crude oil ** Fuures price of Crude oil ** Spo price of gold ** Fuures price of Gold ** Spo price of silver ** Fuures price of Silver ** Spo price of Copper ** Fuures price of Copper ** **Significan a % level From Table 4.2, he regression coefficiens of all he seleced ime series daa are posiive. Hence, he spo prices and fuures prices of Crude Oil, Gold, Silver and Copper show an upward rend. In his chaper, daa analysis was carried ou wih he help of regression model and coinegraion es and inerpreaions were given. The following chaper will cover he summary of findings and recommendaions of he sudy.

44 35 5. CONCLUSION AND RECOMMENDATION The presen sudy was an aemp made o analyze he impac of fuures rading aciviy on he spo prices of Gold, Silver, Copper and Crude Oil and deermine he exisence of relaionship beween fuures prices of Crude Oil, Copper, Silver and Gold. Afer analyzing he available daa on spo prices, fuures prices and fuures rading volume of he seleced commodiies, his chaper provides he summary of research resuls and conclusion for he sudy followed by direcions for fuures research. 5. SUMMARY OF RESEARCH RESULTS The volailiy of spo prices of Gold, Silver, Copper and Crude Oil were deermined year-wise by calculaing he sandard deviaion. I was observed ha spo prices of Silver showed highes volailiy and ha of Copper, he leas. The uni roo es was performed o ascerain he saionariy of ime series daa for seleced commodiies. I was observed ha all he seleced ime series daa were found o be saionary, excep spo prices of Gold and fuures prices of Gold, Silver, Copper and Crude Oil, which were found o be nonsaionary. Muliple regression was used o assess he impac of fuures rading volume on he respecive spo prices of seleced commodiies. From he muliple regression carried ou for spo prices of Crude Oil, i was found ha volume of one monh fuures conrac, volume of wo monhs fuures, volume of hree monhs fuures and volume of four monhs fuures conrac of Crude Oil have a significan impac on spo prices of Crude Oil. ased on he muliple regression for spo prices of Gold, i was observed ha volume of Gold fuures mauring in he firs and second neares monh are having significan impac on spo prices of Gold. Similar resuls were obained for muliple regression on spo prices of Silver, wherein, volume of Silver fuures

45 36 conrac mauring in he firs and second neares monh have significan impac on is spo prices. Ineresingly, he muliple regression for spo prices of Copper showed unique resuls. Volume of Copper fuures conrac mauring in any monh did no have significan impac on spo prices of Copper. This implies ha fuures rading volume of Copper does no influence spo prices of Copper. The relaionship beween fuures prices of Crude Oil, Gold, Silver and Copper were sudied pair-wise using Engle-Granger coinegaion es. The resuls of his es reveals ha, here exiss coinegraion beween closing prices of Crude Oil fuures and Gold fuures, Crude Oil fuures and Silver fuures, Crude Oil fuures and Copper fuures and Silver fuures and Copper fuures. u, here does no exis any coinegraion beween closing prices of Gold fuures and Silver fuures and Gold fuures and Copper fuures. The resuls of rend analysis show ha boh spo and fuures prices of he seleced commodiies show an increasing rend. 5.2 FINANCIAL IMPLICATIONS ased on he resuls obained, i can be observed ha as he conrac mauriy period increases, he relaionship beween he fuures rading volume and spo prices weakens. This shows ha fuures rading aciviy for near-monh fuures conrac has significan impac on he spo prices of he underlying commodiy. This shows ha he very objecive of using commodiy derivaive as a hedging insrumen, o overcome he risk of price flucuaion in he underlying commodiy, is no achieved. Moreover, anoher ineresing finding ha can seen from he sudy is ha, fuures rading aciviy for Copper does no influence he spo prices of Copper. One reason for his can be aribued o he low volailiy in spo prices of Copper from Table 4.2. From his sudy i was found ha fuures rading aciviy has significan impac on he spo prices of he underlying commodiy. I is no only he fuures

46 37 rading aciviy bu also various oher facors which will conribue o price flucuaions in he seleced commodiies. All hese show ha commodiy derivaives marke is dominaed by financial invesors, who ake posiion in he marke, for shor-em gains. Hence seps should be aken o reduce volailiy in prices of commodiies and o make proper use of derivaives for hedging. This can be done by imposing resricions like reducing he open ineres limi, increasing he iniial margin and reducing ick size. A combinaion of derivaive producs like fuures conrac on commodiy, fuures conrac on freigh and Rupee-Dollar forwards can be used by hedgers o reduce exposure. 5.3 DIRECTIONS FOR FUTURE RESEARCH The sudy is limied o four non-agriculural commodiies. I can be exended o include oher commodiies as well. The research carried ou analyzes he impac of fuures rading volume on he spo prices of underlying commodiy. I does no deermine he causal relaionship beween he wo. Hence, o develop on he exising work, Granger causaliy es can be used o find he causal relaionship beween fuures rading volume and spo prices. A case sudy can also be done o bring heory and pracice closer. The Engle-Granger coinegraion es, o deermine he relaionship beween he fuures prices of Crude Oil, Gold, Silver and Copper, can only find he exisence of uni-direcional or bi-direcional coinegraion beween hem. In addiion, Vecor Error Correcion (VECM) can be used o find he coinegraion regression equaion which will be helpful o esimae he fuures price of one commodiy based on he fuures prices of oher commodiies. There are more advanced ess of coinegraion like Johansen Coinegraion ha can be used for his purpose.

47 CONCLUDING REMARKS Indian commodiy derivaive marke has winessed remendous developmens in he pas few years due o increased paricipaion of invesors. As a resul, here is a significan impac on he spo marke which is eviden from he findings of he sudy. These developmens should be made useful o move he commodiy derivaive marke in a posiive direcion for he benefi of he Indian economy.

48 39 REFERENCES [] Corazar, G. and Eerovic, F. (200). Can oil prices help esimae commodiy fuures prices? The cases of copper and silver, Resources Policy, 35, [2] Chng, M. T. (2009). Economic linkages across commodiy fuures: Hedging and rading implicaions, Journal of anking & Finance, 33, [3] Gujarai, D. N. and Sangeeha (2007). asic Economerics (4h Ediion), New Delhi, Taa McGraw-Hill. [4] Ji, Q. and Fan, Y. (202). How does oil price volailiy affec non-energy commodiy markes? Applied Energy, 89, [5 John, C. Hull. (2006). Opions, Fuures and Oher Derivaives (6h Ediion), New York, Prenice Hall. [6] Kumar,. and Pandey, A. (20). Inernaional Linkages of he Indian Commodiy Fuures Marke, Modern Economy, 2, [7] Lokare, S. M. (2007). Commodiy Derivaives and Price Risk Managemen: An Empirical Anecdoe from India, Reserve ank of India Occasional Papers, 28(2), [8] Mukherjee, K. (20). ig ang for commodiies exchanges wais on reform law. [Online] Available:hp:// - wais -on-reform-law_ [9] Nah, Golaka C. and Lingareddy, Tulsi (2008). Commodiy Derivaive Marke and is Impac on Spo Marke. Available a SSRN: hp:// ssrn.com/`absrac= or hp://dx.doi.org/0.239/ssrn [0] Russo, D., Har, T.L., and Schönenberger, A. (2002). The evoluion of clearing and cenral counerpary services for exchange raded derivaives in he Unied Saes and Europe: A comparison. Occasional Paper series, 5, European Cenral ank. [] Sarkar, A. (2006). Indian derivaives marke. In: Kaushik asu (ed.). The Oxford Companion o Economics in India, New Delhi, Oxford Universiy Press. [2] Sieczka, P. and Holys, J. A. (2009). Correlaions in commodiy markes, Physica A 388, [3] Slade, M. E. and Thille, H. (2004). Commodiy Spo Prices: An Exploraory Assessmen of Levels and Volailiies. Journal of Economic Lieraure.

49 40 APPENDIX A. UNIT ROOT TEST The resuls obained for uni roo es on ime series daa are given in he following ables. Uni roo es for variable i is carried ou by regressing variable i on is one-day lagged values. If he regression coefficien is less han one (ρ < ), variable i is saionary. If ρ =, variable i is non-saionary. Table A.: Uni roo es for spo prices of Crude Oil (Consan) Unsandardized Sd. Error Sandardized LAGS (Spo Price of Crude Oil,) Dependen Variable: Spo Price of Crude Oil ea.995 Table A.2: Uni roo es for spo prices of Gold Unsandardized Sandardized Sd. Error (Consan) LAGS(Spo price of Gold,) ea Dependen Variable: Spo Price of Gold Table A.3: Uni roo es for spo prices of Silver Unsandardized Sandardized Sd. Error (Consan) LAGS(Spo Price of Silver,) ea Dependen Variable: Spo Price of Silver Table A.4: Uni roo es for spo prices of Copper Unsandardized Sd. Error LAGS(Spo Price of Copper,).997 Dependen Variable: Spo Price of Copper.003 (Consan) Sandardized ea

50 4 Table A.5: Uni roo es for volume in los of one monh fuures of Crude Oil (Consan) LAGS(Volume of Sandardized Sd. Error ea in los of one monh fuures Unsandardized Crude Oil,) Dependen Variable: Volume in los of one monh fuures of Crude Oil Table A.6: Uni roo es for volume in los of wo monh fuures of Crude Oil (Consan) LAGS(Volume of Sandardized Sd. Error ea in los of wo monh fuures Unsandardized Crude Oil,) Dependen Variable: Volume in los of wo monh fuures of Crude Oil Table A.7: Uni roo es for volume in los of hree monh fuures of Crude Oil (Consan) LAGS(Volume Unsandardized Sandardized Sd. Error ea in los of hree monh fuures of Crude Oil,) Dependen Variable: Volume in los of hree monh fuures of Crude Oil

51 42 Table A.8: Uni roo es for volume in los of four monh fuures of Crude Oil Unsandardized Sandardized (Consan) LAGS(Volume Sd. Error ea in los of four monh fuures of Crude Oil,) Dependen Variable: Volume in los of four monh fuures of Crude Oil Table A.9: Uni roo es for volume in los of five monh fuures of Crude Oil Unsandardized Sandardized (Consan) LAGS(Volume of Sd. Error ea in los of five monh fuures Crude Oil,) Dependen Variable: Volume in los of five monh fuures of Crude Oil Table A.0: Uni roo es for Volume in los of Gold fuures mauring in firs neares monh (Consan) LAGS(Volume in Los of Gold fuures mauring in firs Unsandardized Sandardized Sd. Error ea neares monh,) Dependen Variable: Volume in Los of Gold fuures mauring in firs neares monh

52 43 Table A.: Uni roo es for Volume in los of Gold fuures mauring in second neares monh (Consan) Unsandardized Sandardized Sd. Error ea LAGS(Volume in Los of Gold fuures mauring in second neares monh,) Dependen Variable: Volume in Los of Gold fuures mauring in second neares monh Table A.2: Uni roo es for Volume in los of Gold fuures mauring in hird neares monh (Consan) Unsandardized Sandardized Sd. Error ea LAGS(Volume in Los of Gold fuures mauring in hird neares monh,) Dependen Variable: Volume in Los of Gold fuures mauring in hird neares monh Table A.3: Uni roo es for Volume in los of Silver fuures mauring in firs neares monh Unsandardized Sandardized (Consan) LAGS(Volume in Los Sd. Error ea of Silver fuures mauring in firs neares monh,) Dependen Variable: Volume in Los of Silver fuures mauring in firs neares monh

53 44 Table A.4: Uni roo es for Volume in los of Silver fuures mauring in second neares monh (Consan) Unsandardized Sandardized Sd. Error ea LAGS(Volume in Los of Silver fuures mauring in.707 second neares monh,) Dependen Variable: Volume in Los of Silver fuures mauring in second neares monh Table A.5: Uni roo es for Volume in los of Silver fuures mauring in hird neares monh (Consan) LAGS(Volume in Los Unsandardized Sandardized Sd. Error ea of Silver fuures mauring in hird neares monh,) Dependen Variable: Volume in Los of Silver fuures mauring in hird neares monh Table A.6: Uni roo es for Volume in los of Copper fuures mauring in firs neares monh (Consan) LAGS(Volume in Los Unsandardized Sandardized Sd. Error ea of Copper fuures mauring in firs neares monh,) Dependen Variable: Volume in Los of Copper fuures mauring in firs neares monh

54 45 Table A.7: Uni roo es for Volume in los of Copper fuures mauring in second neares monh (Consan) Unsandardized Sandardized Sd. Error ea LAGS(Volume in Los of Copper fuures mauring in second neares monh,) Dependen Variable: Volume in Los of Copper fuures mauring in second neares monh Table A.8: Uni roo es for Volume in los of Copper fuures mauring in hird neares monh (Consan) Unsandardized Sandardized Sd. Error ea LAGS(Volume in Los of Copper fuures mauring in hird neares monh,) Dependen Variable: Volume in Los of Copper fuures mauring in hird neares monh Table A.9: Uni roo es for Fuures prices of Crude Oil (Consan) LAGS(Closing Unsandardized Sandardized Sd. Error ea Price of one monh fuures.995 of Crude Oil,) Dependen Variable: Closing Price of one monh fuures of Crude Oil

55 46 Table A.20: Uni roo es for Fuures prices of Gold (Consan) Unsandardized Sandardized Sd. Error ea LAGS(Close price of Gold fuures mauring in firs neares monh ) Dependen Variable: Close price of Gold fuures mauring in firs neares monh Table A.2: Uni roo es for Fuures prices of Silver (Consan) Unsandardized Sandardized Sd. Error ea LAGS(Close price of Silver fuures mauring in firs neares monh,) Dependen Variable: Close price of Silver fuures mauring in firs neares monh Table A.22: Uni roo es for Fuures prices of Copper Unsandardized Sandardized (Consan) LAGS(Close price of Copper fuures Sd. Error ea mauring in firs neares monh,) Dependen Variable: Close price of Copper fuures mauring in firs neares monh

56 47 A.2 ENGLE-GRANGER COINTEGRATION TEST In Engle-Granger coinegraion (EG) es, a regression equaion has o be esimaed by regressing he dependen variable on independen variable and heir residuals should be obained. These residuals should be esed for saionariy using Dickey-Fuller es. If he residuals are found o saionary, here exiss coinegraion beween he variables aken for sudy. Oherwise, here does no exis any coinegraion beween he wo variables. The following noaions are used in he ables below: Res_2 - Residuals obained from regression of fuures prices of Crude Oil on fuures prices of Gold Res_4 - Residuals obained from regression of fuures prices of Crude Oil on fuures prices of Silver Res_6 - Residuals obained from regression of fuures prices of Crude Oil on fuures prices of Copper Res_8 - Residuals obained from regression of fuures prices of Gold on fuures prices of Silver Res_0 - Residuals obained from regression of fuures prices of Gold on fuures prices of Copper Res_2 - Residuals obained from regression of fuures prices of Silver on fuures prices of Copper Table A.23: Regression for fuures price of Crude Oil on fuures price of Gold Unsandardized Sd. Error Sandardized ea (Consan) Closing price of Gold fuures mauring in firs neares monh Dependen Variable: Closing Price of one monh fuures of Crude Oil

57 48 Table A.24: EG es for coinegraion from fuures price of Crude Oil o fuures price of Gold Unsandardized Sandardized Sd. Error LAGS(RES_2,) Dependen Variable: DIFF(RES_2,).007 (Consan) ea Table A.25: Regression for fuures price of Crude Oil on fuures price of Silver Unsandardized (Consan) Sd. Error Sandardized ea Closing price of Silver fuures mauring in firs neares monh Dependen Variable: Closing Price of one monh fuures of Crude Oil Table A.26: EG es for coinegraion from fuures price of Crude Oil o fuures price of Silver (Consan) Unsandardized Sandardized Sd. Error LAGS(RES_4,) Dependen Variable: DIFF(RES_4,) ea Table A.27: Regression for fuures price of Crude Oil on fuures price of Copper Unsandardized (Consan) Sd. Error Sandardized ea Closing price of Copper fuures mauring in firs neares monh Dependen Variable: Closing Price of one monh fuures of Crude Oil Table A.28: EG es for coinegraion from fuures price of Crude Oil o fuures price of Copper (Consan) Unsandardized Sandardized Sd. Error LAGS(RES_6,) Dependen Variable: DIFF(RES_6,) ea

58 49 Table A.29: Regression for fuures price of Gold on fuures price of Silver Unsandardized (Consan) Sandardized Sd. Error ea Closing price of Silver fuures mauring in firs neares monh Dependen Variable: Closing price of Gold fuures mauring in firs neares monh Table A.30: EG es for coinegraion from fuures price of Gold o fuures price of Silver (Consan) Unsandardized Sandardized Sd. Error ea LAGS(RES_8,) Dependen Variable: DIFF(RES_8,) Table A.3: Regression for fuures price of Gold on fuures price of Copper Unsandardized (Consan) Sandardized Sd. Error ea Closing price of Copper fuures mauring in firs neares monh Dependen Variable: Closing price of Gold fuures mauring in firs neares monh Table A.32: EG es for coinegraion from fuures price of Gold o fuures price of Copper (Consan) Unsandardized Sandardized Sd. Error LAGS(RES_0,) Dependen Variable: DIFF(RES_0,) ea Table A.33: Regression for fuures price of Silver on fuures price of Copper Unsandardized (Consan) Sd. Error Sandardized ea

59 50 Closing price of Copper fuures mauring in firs neares monh Dependen Variable: Closing price of Silver fuures mauring in firs neares monh Table A.34: EG es for coinegraion from fuures price of Silver o fuures price of Copper Unsandardized Sandardized (Consan) Sd. Error LAGS(RES_2,) Dependen Variable: DIFF(RES_2,) ea

60 5 LIST OF PULICATIONS. Presened a paper iled, An Innovaive Approach o Technical Analysis, in Inernaional Conference on Susainable Innovaion in Global usiness Scenario organized by Madras Chrisian College, Chennai during 4 h and 5h of Sepember Published a chaper iled, Derivaive Usage: Risk Managemen in Essays in Managemen, Excel ooks, New Delhi, pp. 9-3, 202, ISN Published a paper iled, Impac of Commodiy Derivaives Trading on Spo Prices in Conference Proceedings in he 0h AIMS Inernaional Conference on Managemen held a IIM, angalore during January 6 h 9h 203.

61 52 TECHNICAL IOGRAPHY Ms. Kaleel Nisha (RRN ) was born on 23rd May 982, in Neyveli, Tamil Nadu. She did her schooling in Inernaional Indian School, Riyadh, Kingdom of Saudi Arabia and secured 84% in he Higher Secondary Examinaion. She received.sc., Degree in Mahemaics from Theivanai Ammal College for Women, Villupuram in he year 2003 from Madras Universiy, for which, she secured Universiy 5h Rank. She had her Maser s Degree in MA in Finance from SRM Valliammai Engineering College, Anna Universiy, Chennai in he year 2008 and she acquired Anna Universiy 35 h Rank. As of now, she has four years of eaching experience. She is currenly working as Assisan Professor and pursuing her M.Phil. Degree in Managemen in Crescen usiness School of.s. Abdur Rahman Universiy. The ID is : kaleelnisha@bsauniv.ac.in.