EURASIAN JOURNAL OF ECONOMICS AND FINANCE

Transcription

1 Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 DOI: /ejef EURASIAN JOURNAL OF ECONOMICS AND FINANCE hp:// TRADING ACTIVITY AND PRICES IN ENERGY FUTURES MARKET Aysegul Aes Akdeniz Universiy, Turkey. Absrac This paper aims o examine rading aciviy and he relaionship beween fuures rading aciviy by rader ype and energy price movemens in hree energy fuures markes naural gas, crude oil and heaing oil. We find ha he level of ne posiions of speculaors are posiively relaed o fuure reurns and in conras ne posiions of hedgers are negaively relaed o fuures price changes in all hree markes. The changes in ne posiions are relaively more informaive compare o he level of ne posiions in predicing price changes in relaed markes. Keywords: Energy Fuures, Speculaion, Hedging Pressure, Trading Aciviy 1. Inroducion Sharp price movemens of energy prices in he mid-2s led o concerns by policy makers and regulaors. The rading of energy derivaives conracs have begun o grow rapidly over he pas decade compare o cash markes (Domanski and Heah, 27). The coincidence of increased derivaives rading wih increasing energy prices has driven sudies which examine wheher price effecs have been augmened by his developmen in energy markes. Some sudies such as Tang and Xiong (21) argued ha increasing paricipaion of financial invesors in he fuures marke rading (i.e., invesmens in commodiy indices) were he large par of he commodiy price movemens and commodiy fuures prices are driven by speculaive pressures. 1,2,3 On he oher hand hedging pressure heory (Keynes, 193) saes ha hedgers who are ypically ne shor on he fuures markes offer a risk premium o arac speculaors o ake a long posiion and provide liquidiy o hedgers and his migh resul in changed price levels. 4 In sum, boh hedging and speculaive pressures assume ha changes in fuures prices are affeced by changes in hedgers' and speculaors' open ineres posiions. In his paper by employing CFTC s Commimen of Trader (COT) repor, we will examine he relaion beween fuures rading aciviy by rader ype and price changes for seleced energy markes and ry o answer wheher fuures rading aciviies convey valuable 1 Faouh e al. (212) provides an exensive survey of he role of he speculaion in oil markes. 2 For example, energy commodiies accouned for abou one-hird of 215 arge weighs in Bloomberg Commodiy Index wih crude oil comprising 15%. 3 Please noe ha here are oher sudies such as Krugman (28), Kilian (29) and Wuhisaian (214) ec. ha sugges price movemens are originaed from fundamenals raher han excessive speculaion. 4 Various sudies have documened he effec of hedging pressure on prices such as Bessembinder (1992) and De Roon e al. (2).

2 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 informaion on poenial price movemens on energy markes. CFTC s COT repor publishes informaion on he posiions held by reporing raders. COT repor disaggregaes open ineres ino posiions held by reporing and non-reporing raders and former caegory furher disaggregaed as commercials and non-commercials. Tradiionally commercial caegory is viewed as commercial hedgers and non-commercial caegory as speculaors. 5 We use ne posiions of raders as a proxy for rading aciviy. Examinaion of rading aciviy (i.e. boh speculaion and hedging aciviy) could help us o deermine wheher here is an excessive speculaion in he US energy fuures and he price impac of fuures rading aciviy. Undersanding he facors behind hese price movemens in energy markes will be useful for policy makers and regulaors as well as he raders of hese markes. This paper examines he price impac of fuures rading aciviy (boh speculaion and hedging) on energy prices. The remainder of his aricle is organized as follows. Secion 2 explains he mehodology employed in he paper. Secion 3 explains he daa and presens preliminary analysis of he daa. Our resuls are presened in Secion 4 and Secion 5 concludes he paper. 2. Mehodology Firsly we will invesigae how raders' curren posiions and he changes in raders' ne posiion impac he fuures price movemens. In order o do so, following regressions would be esimaed p NP, (1) k j and p NP, (2) k j where p 1 p p k k is k period ahead log price changes, k=,1,4,8,12. For k= p p p is he curren change in he log prices. NP is he level of ne posiions (speculaors or hedgers) and is he change (in level) over previous j NP, j NP NP, j period's ne posiion. In his framework he ne of raders' long posiions versus heir oal shor posiions is inerpreed as a measure of heir rading aciviy (i.e., speculaive or hedging aciviies) in he fuures markes. The ne posiions held by hedgers indicaed hedging demand. We also calculae hedging pressure 6 in he markes using following formula in order o deermine he exen of hedging pressure in energy markes: Hedging pressure= Shor commercial posiions-long commercial posiions/toal commercial posiions (3) Secondly we will calculae Working's T index which measures he adequacy of speculaive posiions needed o balance he hedging posiions held by commercial raders. The T index has a minimum value of 1 when speculaion level is equal o he hedging need. According o Working (196), he speculaion aciviy which exceeds hedging needs is considerae as excessive and excessive speculaion could cause price volailiy in relaed markes. Working's speculaive T index is calculaed as follows: 5 We mus noe ha by 29 CFTC launched a more deailed repor of raders called as Disaggregaed Commimens of Traders (DCOT) which adds four new caegories of large raders (Producer/merchan/processor/user, swap dealers, managed money, oher reporables). One drawback of his daa se is ha i was no available during he surge of commodiy fuures rading and emergency of new financial paricipans during he mid 2s. 6 There are sudies which examine he impac of hedging pressure on he marke prices. See Bessembinder (1992), Tien (22), and DeRoon e al. (2) ec. 2

3 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 T =1+SS/(HL+ HS) if HS HL (4) or T = 1+SL/(HL + HS) if HL HS (5) where open ineres held by speculaors (non commercials) and hedgers (commercials) is denoed as follows: SS is speculaion shor, SL is speculaion long, HS is hedging shor and HL is hedging long. 3. Daa and Preliminary Analysis Daa relaed o naural gas, crude oil and heaing oil fuures markes are obained from U.S. Commodiy Fuures Trading Commission (CFTC). The nearby fuures conracs are employed in his sudy. Selemen prices and open ineres for all conracs are from CFTC daabase. Conracs used are crude oil (CL), naural gas (NG), heaing oil (HO) for he period Ocober 6h 1992 o December 13h 211. Hisorical daa on he fuures rader posiions, used in our analysis are exraced from he Commimens of Traders (COT) repors. The fuures marke open ineres posiions of marke paricipans are colleced every Tuesday and his rader posiion informaion released weekly o he public on following Friday since Ocober The repors conains informaion on he long and shor posiions of he hree caegories (Reporable (Commercials, Non-Commercials) and Non-Reporable) as well as open ineres across all mauriies of fuures conracs. In each marke, he CFTC defines large raders. If a rader s posiion exceeds he reporing hreshold hen he/she is classified as large rader. Reporing levels are adjused periodically by CFTC. Aggregae posiions in Large Trader Reporing accoun for 7 or 9 percen of open ineres in any given marke. 7 The legacy COT repor separaes reporable raders only ino commercial and non-commercial caegories. A rader is classified as a commercial rader if rader uses derivaives for hedging agains commodiy price risks in heir businesses. If a rader akes fuures posiions for reasons oher han hedging, he rader is regarded as noncommercial. Following he lieraure we inerpre he commercial raders as commercial hedgers (who were involved handling he physical commodiy) and noncommercial raders as speculaors. Figure 1 presens nearby fuures prices for he daa period. Descripive saisics and ADF Augmened Dickey Fuller (ADF) ess for he fuures price changes and he ne posiions held by speculaors and hedgers (i.e., rading aciviies) for naural gas, crude oil and heaing oil fuures markes by rader ype are repored in Table 1. ADF es resuls sugges prices I(1) and ne posiions end o be saionary (i.e., I()) wih he excepion of naural gas markes. The means of hedgers' ne posiions found o be negaive for crude oil and heaing oil marke and posiive for naural gas markes. Tha means on average hedgers are ne shor on crude oil and heaing oil markes and hey are ne long on naural gas markes. Ineresingly, ne open posiions of speculaors are larger han he ne open posiions of hedgers in naural gas marke on average. On he oher hand, ne open posiions of hedgers and speculaors are roughly same in crude oil marke. However, large sandard deviaions in naural gas and crude oil markes indicae ha subsanial variaions in ne posiions. The average weekly price changes are posiive for all hree markes. This suggess susained period of upward movemen in seleced energy fuures markes. 7 Classificaion of commercials ino hedgers, non-commercials ino speculaors and non-reporables ino small raders is done by CFTC. Trader designaions may be inaccurae. Furhermore, his classificaion migh be misleading wih he growh of commodiy index invesing 3

4 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 Naural gas fuures nearby conrac prices Crude oil nearby fuures price Heaing oil nearby fuures Figure 1. Nearby fuures prices 4

5 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 Table 1. Summary saisics Naural gas Crude oil Heaing oil Ne Posiions Hedger Mean of NP Median of NP S. dev of NP ADF Tes of NP Mean of NP * *.1 Mean of NP Ne Posiions Speculaor Mean of NP Median of NP S. dev of NP ADF Tes of NP Mean of NP * * -.4 Mean of NP Price Mean Median S. dev ADF Tes * * * Noes: Ne posiions (NPs) are measured in uni of 1, conracs. * denoes significance level a 1%. 4. Empirical Resuls Figure 2 presens hedging posiions in 1 conracs. On average hedgers ake ne shor posiions and speculaors ake ne long posiions for all markes wih he excepion of naural gas marke. No surprisingly hedgers and speculaors ake opposie posiions. Hedgers hold he larges ne shor posiions in crude oil marke and posiions of speculaors and hedgers roughly offse each oher in crude oil marke. Our resuls for naural gas fuures markes are in conras wih Sanders e al. (24) who find for energy markes commercials are ne shor and non commercials are ne long, deparure from heir resuls sem from he use of he exended and more recen daa se which includes he subsanial increase in rading aciviy in commodiy fuures conracs whereas Sanders' daa se ends in December 28, This resul may also indicae ha here was a srucural change in naural gas rading afer he mid 2s. Figure 3 presens hedging pressure resuls. 8 Hedging pressure values presens mixed paerns wih some posiive some negaive values indicaing a range of differen hedging behavior across markes and ime. In naural gas marke here is noiceable increase in percen of long posiions afer 25. The resuls of Working Index is presened on Figure 4. As i can be seen from Figure 4 here is a noiceable upward in he index for boh crude oil and naural gas fuures markes saring from he beginning of Hedging pressure= Shor commercial posiions - Long commercial posiions / Toal commercial posiions. 5

6 5/22/1 5/14/2 4/29/3 3/29/5 3/14/6 2/27/7 1/27/9 1/12/1 12/28/1 5/22/1 5/14/2 4/29/3 3/29/5 3/14/6 2/27/7 1/27/9 1/12/1 12/28/1 5/22/1 5/14/2 4/29/3 3/29/5 3/14/6 2/27/7 1/27/9 1/12/1 12/28/1 5/22/1 5/14/2 4/29/3 3/29/5 3/14/6 2/27/7 1/27/9 1/12/1 12/28/1 5/22/1 5/14/2 4/29/3 3/29/5 3/14/6 2/27/7 1/27/9 1/12/1 12/28/1 5/22/1 5/14/2 4/29/3 3/29/5 3/14/6 2/27/7 1/27/9 1/12/1 12/28/1 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, Naural gas Hedging Ne Posiions (in 1 conracs) 1 Crude Oil Hedging ne posiions (in 1 conracs) Naural gas Speculaive Ne Posiions (in 1 conracs) Crude Oil Speculaive ne posiions (in 1 conracs) Heaing Oil Hedging Posiions (in 1 conracs) Heaing Oil Speculaive Posiions (in 1 conracs) Figure 2. Ne posiions 6

7 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 Naural gas Hedging Pressure=ShorCommercial-LongCommercials/Toal Commercials Crude Oil Hedging Pressure=ShorCommercial-LongCommercials/Toal Commercials Heaing Oil Hedging Pressure=ShorCommercial-LongCommercials/Toal Commercials Figure 3. Hedging pressure 7

8 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, Naural Gas Workings T index Crude oil Workings T Index Heaing Oil Workings T index Figure 4. Working s T index Lasly we examine he effecs of ne posiions and changes in ne posiions on energy fuures price changes, he resuls are given Tables 2 and 3 respecively. 8

9 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 Table 2. Ne posiions and energy price regressions Naural gas Crude oil Heaing oil Hedger Speculaor Hedger Speculaor Hedger Speculaor k= β.19 (.7947) -.13* ( ).38 (1.4193).* (2.7152) -.16 (-.935) -.11* (-4.268) -.2 ( ).13* (4.222) -.12* ( ) -.61* (-4.349) -.39 ( ).71* (3.5953) k=1 β.19 (.827) -.13* ( ).39 (1.4828).* (2.8589).7 (.4165) -.2 (-1.51).6 (.3187).3 (.9665).5 (.1238) -.6 (-.3555).11 (.3655).6 (.2834) k=4 β.52 (1.113) -.32* ( ).17 (2.63).* (3.8116).44 (1.3331) -.6 ( ).48 (1.428).4 (.7991).61 (1.1312) -. (-.163).53 (1.2765).112 (.4232) k=8 β.75 (1.1332) -.4* (-3.714).153 (2.83).* (3.5595).96* (2.683) -.9 ( ).12* (2.123).7 (.91).92 (1.346) -.19 (-.7226).95** (1.847).46 (1.349) k=12 β.12 (1.2912) -.46* ( ).21 (2.353).* (3.9562).136* (2.3257) -.16* ( ).145* (2.3963).13 (1.2822).127 (1.5574) -.36 (-1.166).139* (2.2195).74** (1.7755) Noes: p +k=+βnp +ε. -saisics are given in parenhesis. *, ** and *** represen 1%, 5% and 1% significance level respecively. The conemporaneous relaionship presened in firs row is significan for all markes for Ne Posiion regression presened a Table 2. This indicaes ha ne posiions and prices move ogeher. The esimaed slope coefficiens are uniformly negaive for hedgers and posiive for speculaors hroughou all conracs. However, he evidence sugges ha predicive power he ne posiions is fairly weak for all markes excep naural gas fuures marke. 9 For crude oil marke only hedgers' ne posiions are predicor of changes in prices over he 12 week horizon and for heaing oil marke only speculaors' ne posiion are significan over he 12 week horizon. In sum, he speculaors' and hedgers' ne posiions mosly do no have predicive power over he relevan horizons for he examined conracs. In general he esimaed inerceps for fuure price changes are insignifican. Table 3 presens regression resuls peraining o he relaionship beween changes in ne posiions and energy price movemens. We examine he effec of curren weekly changes of ne posiions on fuures price changes. The changes in ne posiions are relaively more informaive compare o he level of ne posiions. Specifically curren changes in ne posiion can forecas fuures price changes up o 12 weeks ahead for naural gas marke and up o 4 weeks in crude oil markes. For heaing oil marke he coefficien of changes in ne posiions are significan up o 1 week ahead. I is also ineresing o noe ha changes in hedgers' ne posiions end o negaively covary wih fuures price changes. This negaive co-variaion sugges ha hedgers hold more long conracs on he fuures marke before he fuures prices move downward. 9 However, he resuls relaed o naural gas markes should be inerpreed wih cauion since ne posiion daa are no saionary. 9

10 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 Table 3. Change of ne posiions and energy price regressions Naural gas Crude oil Heaing oil Hedger Speculaor Hedger Speculaor Hedger Speculaor k= β.4 (.1683).13 (.4877).3 (.1495) -.38 ( ).13 (.8581) -.122* ( ).13 (.864).141* (13.434).15 (.6331) * ( ).15 (.6265).6461* ( ) k=1 β.5 (.2282) -.76* ( ).5 (.2351).79* (2.529).15 (.93) -.25* (-2.53).15 (.98).32* (2.7789).15 (.25) -.814** (-2.484).15 (.5925).72 (1.3715) k=4 β.25 (.557) -.533* (-.5687).27 (.616).586* (1.36).59** (1.9761) -.35** ( ).59** (1.9736).41** (1.8843).61* (1.7281) -.57 (-1.379).61** (1.7286).383 (.5253) k=8 β.42 (.6596) -.559* ( ).44 (.6946).635* (7.5232).122* (2.845) -.19 (-.7424).122* (2.8453).16 (.539).129* (2.8485).14 (.251).129* (2.8489) -.43 (-.4611) k=12 β.66 (.8639) -.585* ( ).69 (.961).662* (6.5314).181* (3.3424). (.213).18* (3.3391).9 (.2475).196* (3.62) -.36 (-.4224).196* (3.62).28 (.2493) Noes: p +k=+β NP +ε. -saisics are given in parenhesis. *, ** and *** represen 1%, 5% and 1% significance level respecively. 6. Conclusion We presen an empirical analysis in his paper by invesigaing he relaionship beween fuures rading aciviy and changes in energy fuures prices in seleced energy markes. In conclusion his paper provides evidence ha he level of ne posiions mainly is no helpful explaining price changes in examined markes wih he excepion of conemporaneous esimaions. Our resuls sugges ha change in he ne posiions raher han he level of ne posiions has an impac on fuures price movemens. Thus he policy argumen aimed a limiing speculaive posiions in energy markes are no suppored by empirical evidence. References Bessembinder, H., Sysemaic risk, hedging pressure and risk premiums in fuures markes. Review of Financial Sudies, 5, pp hp://dx.doi.org/1.193/rfs/ DeRoon, F., Nijman, T., and Veld, C. 2. Hedging pressure effecs in fuures markes. Journal of Finance, 55, pp hp://dx.doi.org/1.1111/ Domanski, D. and Heah, A., 27. Financial invesors and commodiy markes. Bank for Inernaional Selemens Quarerly Review, March(27), pp Faouh, B., Kilian, L., and Mahadeva, L., 212. The role of speculaion on oil markes: Wha we have learned so far?. CEPR Discussion Paper. London: Cenre for Economic Policy Research. 1

11 Aysegul Aes / Eurasian Journal of Economics and Finance, 4(2), 216, 1-11 Keynes, J., 193. Treaise on money. London: Macmillian. Kilian, L., 29. No all oil price shocks are alike: Disenangling demand and supply shocks in he crude oil marke. American Economic Review, 99, pp hp://dx.doi.org/1.1257/aer Krugman, P., 28. More on oil and speculaion. The New York Times, [blog] 13 May. Available a:<hp://krugman.blogs.nyimes.com/28/5/13/more-on-oil-and-speculaion/> [Accessed 17 April 214]. Sanders, D.R., Boris, K., and Manfredo, M., 24. Hedge funds and small speculaors in he energy fuures markes: An analysis of he CFTC's commimens of raders repor. Energy Economics, 26, pp hp://dx.doi.org/1.116/j.eneco Tang, K. and Xiong, W., 212. Index invesing and he financializaion of commodiies. Financial Analyss Journal, 68(6), pp hp://dx.doi.org/1.2469/faj.v68.n6.5 Tien, D., 22. Hedging demand and foreign exchange risk premia. Working Paper, Hass School of Business, UC Berkeley. Working, H., 196. Speculaion on hedging markes. Food Research Insiue Sudies,1(2), pp Wuhisaian, R., 214. Governmen resource subsidy and is spillover effecs: Evidence from he excessive oil consumpion in China. Eurasian Journal of Economics and Finance, 2(1), pp