Loading...
HomeMy WebLinkAboutAPA3329C. 6 [ E C [ [ c r L [ [ c c c c p u [j c [ . '"'"' .vc .--,.....,'I ·q t \...-.) .) L.. 4 Energy Di v is1 on £nergy and Economic Analysis Section f~e~.tt. lO ORNL/Tt+-9521 FUlURE WORD OIL PRICES: MODELING ME1HOOOLOOIES AND SUMMARY OF RECENT FORECASTS T. Randall Curlee Date Publ fshed:' March 1985 Prepared by the O.AJ< RICGE NATION~\. L,4BORATORY Oak Ridge, Tennessee 37831 operated by MARTIN MARIETTA ENEFGY SYSTEMS, INC. for the U.S. DEPARTMENT OF ENERGY under Contract No. DE-ACOS-840R21400 ] J l n d ] I J J J J J J J J ] J ] J J n [ [ [ r '-~ [ L c r L [ [ c c [ c r L [ [ [ TABLE OF CONTENTS 4 . . . . . . . . LIST OF FIGURES • LIST OF T,IBLES . . . . .. . . . . . . . . . . iv . . . . . . . v MSTRACT • . . . . . . . . . . . . . . . vi Chapter 1: INTRODUCTION . . . . . . . . . . . 1 Chapter 2: METHODOLOGIES TO EXPLAIN WORLD OIL PRICE BEHAVIOR • 5 I. Introduction • • • • • • • • • • • • • • • • • 5 II. A Brief Overview of Historical Changes in the World 011 Market • • • • • • • • • • • • • • • 5 A. The 196Q-73 Time Period • • • • • • • • • 6 B. The 1973-81 Time Period • • • • • • • • • 7 C. The 1981 to Present Time Period • • • • • 11 III. Methodologies Developed to Explain and Forecast Oil Prices • • • • • • • • • • • • • • • • • • 15 A. Economic Models • • • • • • • • • • • • • 18 1. Optimization Models • • • • • • • • • 18 2. Simulation Models •••••••••• 23 B. Political Models • • • • • • • • • • • • • 25 C. Combination Models • • • • • • • • • • • • 27 IV. Conclusions ••••••••••••••••• 28 Chapter 3: WORLD OIL PRICE FORECASTS • • • • • • • • • • • • • 31 I. Introduction • • • • • • • • • • • • • • • • • 31 II. Forecasts Based on Expert Judgment • • • • • • 32 III. Forecasts Based on Formal Models ••••••• 36 A. The Forecasts • • • • • • • • • • • • • • 37 B. Sensitivities of Forecasts to Changes in Key Parameters • • • • • • • • • • • • • • SO 1. Long-Run Oil Price Elasticity of Demand 53 2. GNP Growth • • • • ·• • • • • • • • • • 54 3. Oil Supply Price Elasticity ••••• 55 4. Interest Rates • • • • • • • • • . • • 57 5. The Backstop Price ••••.••••• 57 IV. Surveys of 011 Price Forecasts • . • • • • . . 58 Chapter 4: CONCLUSIONS • • • • • • • • • • • • • • • • • • • • 61 REFERENCES I. Introduction • • • • • • • • • • • • • • • • • 61 II. A Summary of Conclusions from Previous Chapters 62 III. A Subjective Assessment of Future Market Trends 65 IV. Future Oil Prices .••••.•••.•••• 70 V. Final Conclusions •.•.••••.••••• 74 . . . . . . . . 75 i 11 c [ [ r [ [ [ c r L [ [ c c E [ f' u [ L L LIST OF FIGURES .. Figure 2.1 Methodolog~es to Forecast World Oil Prices • • • • • • • • 17 3.1 4.1 Sensitivities of EMF-6 Price Forecasts •• 52 EMF-6 Price Projections Given a Permanent 10 Million Barrel Per Day Capacity Reduction • • • • • • • • • • • • • • • • 73 1v c c [ E [ [ [ c r L [ [ u 0 E c f' u l f' w [ LIST OF TPSLES 4 Table 2.1 Official Prices of Saudi Arabian Light and Nigerian Crude for Se 1 ected Dates • • • • • • • • • • • • • • • • • • • • 16 3.1 A Summary of Oil Price Forecasts Arranged by Year Published 38 3.2 A Summary of Oil Price Forecasts Arranged by Year Published <In constant 1981 dollars) •.•••••••••••••• 40 3.3 A Summary of Oil Price Forecasts Arranged by Modeling Type 42 3.4 A Summary of Oil Price Forecasts Arranged by Organization Making Forecast • • • • • • • • • • • • • • • • • • • • • 44 3.5 A Summary of Oil Price Forecasts Arranged by Projected 1990 Price Level . • • . . . . • . • •.•. , • • • • • • • • 46 3.6 Scenario Assumptions Used in.the EMf-6 Study. 3.7 OECD Oil Supply Elasticities: EMF-6 Models . . . . 3.8 Results of the 1983 and 1984 IEW Surveys . 4.1 011 Price ProJections: 1987 to 2022 •.• v 51 56 60 71 c [ [ E [ [ [ 0 r L, [ [ c [J E c r u [ [ [ 4 FUTURE WORLD OIL PRICES: MODELING METHOOOLCGIES AND SUMMARY OF RECENT FORECASTS T. Randall Curlee Energy and Economic Analysis Section Energy Division Oak Ridge National Laboratory ABSTRACT This paper has three main objectives. First, the various methodologies that have been developed to explain historical oil price changes and forecast future price trends are reviewed and summarized. Second, the paper summarizes recent world oil price forecasts, and, when possible, discusses the methodologies used in formulating those forecasts. Third, utilizing conclusions from the reviews of the modeling methodologies and the recent price forecasts, in combination with an assessment of recent and projected oil market trends, oil price projections are given for the time period 1987 to 2022. The paper argues that modeling methodologies have undergone significant evolution during the past decade as modelers increasingly recog~ize the complex and constantly changing structure of the world oil market. Unfortunately, at this point in time a· consensus about the appropriate methodology to use in formulating oil price forecasts is yet to be reached. There is, however, a general movement toward the opinion that both economic and political factors should be considered when making price projections. Likewise, there is no consensus about future oil price trends. Forecasts differ widely. However, in general, forecasts have been adjusted downwardly in recent years. Further, an overall assessment of the forecasts and recent oil market trends suggests that oil prices will remain constant in real terms for the remainder of the 1980s. Real oil prices are expected to increase by between 2 and 3 percent during the 1990s and beyond. Forecasters are quick to point out, however, that all forecasts are subject to significant uncertainty. vi C c [ [ [ r~ L c r b [ [ c C E c r L l L [ .. 1. INTRODUCTION Between the years 1973 and 1982 the nominal price of oil on the world market Increased by more than 1600 percent--from $2.10 to $34.00 per barre1.1 Further, the price Increases were sudden and sharp, Increasing from $2.10 to $9.60 per barrel between January 1973 and January 1974 and from $13.34 to $26.00 per barrel between January 1979 to January 1980. These price escalations--both In terms of their stze and their occurrence over a relatively short time period have caused. and continue to cause, significant Impacts on economic activity, social structures. government programs, and Jnternatfonal relations. These price escalations and their resulting Impacts have also spawned a host of studies that espouse particular methodologies for explaining historical price changes, as wei I as for forecasting future price trends. Yet more than ten years after the start of the so called "Energy Crisis" there exists I lttle consensus about how the world oil market functions and, more speclflcal ly, how world oil prices are de term I ned. This Is not an·unexpected outcome, given the accuracy of prevIous forecasts. Further, the poor accuracy of previous forecasts Is not altogether surprising, given that the model lng assumptions used In those forecasts did not reflect the unexpected and what most would argue were unpredictable --market disruptions ot 1 973-74 and 1979-80. Work continues to study the structure of the world oil market and the mechanism by which oil 1 Price quotes are for Saudi Arabia I lght crude, the official marker crude for the Organization of Petroleum Exporting Countries <OPEC>. c c [ [ [ [ [ 0 r L [ [ [J c E c r L [ 6 [ 2 prices are determined. Work also continues to forecast oil prices using models with wJdely divergent underlying assumptions. The purpose~ of this· paper are threefold. First, the paper rev Jews and to explain summartzes the various methodologies that have evolved historical oil price changes and to forecast future price trends. Second, the paper summarizes the numerous world oil price forecasts that have appeared in recent years and Identifies, when possible, the methodologies and underlying assumptions used in formulating those forecasts. Third, utilizing the conclusions drawn from the reviews of the model Tng methodologies and the recent price forecasts, In combination with an projected oil market trends, oil price the time period 1987 to 2022. assessment of recent and projections are given for Since the oil price shock of 1973, numerous theories and models have been developed that supposedly represent the functioning of the world oil market. In Chapter 2 these various 'views of the oil market are reviewed and categorized into three major groups -- economic models, political models, and models that combine aspects of the two approaches. Under each of these three main headings there Is additional dJsaggregatJon to reflect different objective functions, structural assumptions about the world of I market, and so forth. Gtven the large number of oil models, It Is not feasible nor necessarily productive --to attempt to review each Individual model. Rather, this review points out commonal Itles among the models and methodologies with the purposes of (1) suggesting the relevant players that determine the price of world of I, <2> suggesting the objective functions of those players, and c c [ C [ [ [ 0 r 6 [ [ 0 ~ 6 c r u u L b [ 3 (3) s~ggesttng the degree to which each of the players Is able to manipulate the market to maximize or satisfy their objectives. Another lmporta~t aspect·of this review Is the Identification of the major parameters that have Impacted, and can be expected to Impact, world oil prices under the different methodologies. As wfl I become apparent. there are numerous ways that the world of I market can be modeled. Further, each of those approaches have certain credlbl lltles --as wei I as handicaps--In explaining historical oil price changes. However, different methodologies m~_ suggest very different reasons why oil prices m~ have changed In the past and why prices may change In the future. Depending upon how one characterizes the world oil market, the lmpl lcatlons for how future oil prices may change are quite different. In Chapter 3 the numerous oJI price forecasts that have been publ Jshed Jn recent years are reviewed and summarized. In addition to reporting how different forecasters view future oil price trends, the chapter Identifies, where possible, the model Jng methodologies and assumptions about key parameters used to obtain the forecasts. Of particular Importance are the sensitivities of the different forecasts to changes In major Input parameters. An additional point of Interest Is how forecasts of mid-term and long- term oil prices have changed since the official oil price reduction of 1983. In the final chapter the conclusions from Chapters 2 and 3 are used In combination with an assessment of current and projected oil market conditions to select high, medium, and low oil price projections for the tlme perlcd 1987 to 2022. In addition to c [ [ c C [ [ c r L [ [ c [J E c f' u ~~ c L [ 4 selecting business-as-usual price paths --I.e., paths for which the baste structure of the world oil market remains unchanged-- two price path~ represen-ting the breakup of the oil cartel and the occurrence of a significant oil supply disruption are discussed. ' c c [ E [ [ [ c r L [ [ c E E G r L L 6 [ 5 • 2. METHOOOLOOIES TO EXPLAIN WORLD OIL PR ICE 8 EHAV I OR I. lntrocluctto~ Thts chapter reviews and summarizes the various methodologies that have been developed to explain historical world oil price changes and forecast future oil price trends. The methodologies are categorized Into one of three broad groups--(1) economic, (2) pol tttcal, and <3> economic-pol tttcal combinations. In addition to explaining how the different methodologies view the relevant players In the world otl market and the objective functions of those players (If any>, the review Identifies the major parameters relevant to each major methodology that are argued to have Impacted, and are expected to Impact, world oil prices. However, before those methodologies are addressed In detail, a brief historical review of major developments In the world oil market Is necessary to understand why particular methodologies have evolved. That review Is presented In the next section. II. A Brlef Overview of Hlstorlcal Changes ln the World Oll Market For our purposes the post-OPEC era of the world oll market can be divided Into three tlme periods--1960-73, 1973-81, and 1981 to the present. An examination of each of these periods spel Is significant changes not only ln the levels of world of I prlces, but also for the relevant players and market structural condltlons under which those players acted Individually or as a group to determine or at least Impact world oil prices. c c [ E [ [ [ 6 r L [ [ c E E c r L l [ [ 6 A. Tne 196Q-73 TT me Pert od The Organization of Petroleum Exporting Countries <OPEC> was formed In 1960 (lnd consisted of five members--Iran, Iraq, Kuwatt, Venezuela, and Saudi Arabta. At the time of the formation of OPEC, the oil producing nations wfthin OPEC, and the other oil producers that would later jofn the organtzatton, had I ittle power over production decrstons wlthfn thefr own nations, and of course even less Influence on world oil prtces. During the 1950s and most of the 1960s the production and trade of ofl from the <to become> OPEC countries was control led to a great extent by the major internatfo~al oil companies. World otl prices were generally below the $2.00 per barrel level and international oil companies held a very high equity Interest In the crude production of the major OPEC producers. The taxes received by the producing countries from the production of a barrel of oil were very low. The objectives of OPEC tn its early years were rather unambitious as compared to later years. The producing countries attempted to (1) gain control over the level of production in their own countries, (2) change the tax system by which the producing countries received revenues from the major of I companies so as to Increase the total revenues to the producing countries, and (3) gain greater equity interest In the production operations tn their own countries. However, to a great extent the productng countries did not begin to effectively use their tnternational power until around 1970. For example, In January 1970 the new radical Libyan government forced International oil --or threatened with national tzatlon --the companies operating within Libya to renegotiate D [ [ D [ [ [ c r L [ [ c E B c r L t~ L l 7 their oil prices. This event was probably most Important In that It provided a raJ lying point for other oft producers and brought the real Jzatlon_that the ·producing countries had considerable power to Impact the oil market. Later In 1970 both Algeria and Iraq won concessions from the oil companies over prices and Investment plans. These events seemed to set off a series of successful price and benefit demands made by the producing countries. The major point to note from the review of this period Is that the'major oft producing countries gained signlfJcant power In determining both the price and productlon of oil In their own countries and therefore became a new and powerful player In the determination of world oil prices. In other words. there was a change In (1} the relevant players In the determJnatlon of oil prices. <2> the feasible objectives of those players. and (3) the abl I ltles of the different players to manipulate the market. or at least their segments of the market. to real lze those objectives. Although the feasible objectives of OPEC remained I Jmlted -- compared to future standards --the organization had become a power to be dealt with In the world oil market. B. The 1973-81 Time Period Between the years 1973 to 1981 the world oil market changed tremendous I y. Previous to the major events of 1973-74 the producing countries had attempted to obtain a larger share of the economic rents that had gone to either the International oil companies or to the treasuries of the consuming countries ln the form of taxes. (Prlces of petroleum products were much higher during the 1960s and early 1970s than could be accounted for by c [ [ t [ [ [ c r Lo [ [ c E E c r '=i l c L 8 crude 4 prlces. For example, prfces for gasol lne In Europe were In the $30.00 per barrel price range reflecting high domestic tax levels.) Afte~ 1973 the·produclng countries not only continued to transfer the wealth that was previously obtained by the oil companies and the governments obtained additional wealth by crude to consuming countries. of consuming countries, directly Increasing the but also price of The first major price jolt of the 1970s came In October 1973 when six major Persian Gulf producers --Abu Dhabi, Iran, Iraq, Kuwait, Qatar, and Saudi Arabia--agreed to raise the posted price of Saudi marker crude from $3.01 to $5.12 per barrel. Then In January 1974 the Organization of Arab Petroleum Exporting Countries COAPEC) further raised the posted price of Saudi marker crude to $1 1 .65 per barrel, ref I ectl ng more than a 380 percent I ncr ease t n posted prices over the pre-October level. These huge price Increases, of course, occurred during the time of the Arab oil embargo against oil sales to the United States and the Netherlands because of their support of Israel In the 1973 Arab-Israeli War. However, close Investigation of that embargo shows that although approximately 98 percent of the ol I from OAPEC producers was halted to the U.S., Increased Imports from other sources helped to mitigate the Impacts of the embargo. One factor that helped reduce the Impacts of the embargo on physical flows of oil was the fairly good control that the major oil companies retained over the distribution of world o! I. While OAPEC oil co~ld not be sent to embargoed countries (because of threats from the ol I producing p LJ c [ c [ [ [ c c L [ [ 0 D E c r u [~ L r' ld [ 9 counttles>, other oil could easl ly be rerouted to the U.S. and the Nether I ands. Sfmultaneou~ly, several producers reduced production In response to the embargo. In October 1973 Saudf Arabia, Kuwaft, Dubaf, and Algeria cut production by over 5 percent. These same countries cut production by a minimum of 25 percent In November 1973, and both Abu Dhabi and Qatar came very close to these targets during the same tfme period. However, production cutbacks began to diminish in December. By January 1974 production cutbacks in these countries were only about 10 percent. Further, it Is Interesting to note that not alI OPEC countries participated in the production cutbacks and some even Increased production. For example, Libya, a major supporter of the embargo, met the OAPEC mandated 25 percent reduction only tn November. Production actually Increased by about 4 percent In Iran, Indonesia, and Nigeria. Therefore, the rather moderate production cutback was maintained for only about three months and was not shared equally by alI OPEC members. The period 1975 to 1978 was relatively tranqul I as the world attempted to adjust to the huge price Increases of 1973 and 1974. In nominal terms world oil prices continued to Increase at between 5 and 10 percent per year, while In real terms the price of ol I actually decl lned. The second major jolt to world ol I prices followed the start of the Iranian Revolution In late 1978. Oil exports from Iran were reduced In September 1978 and by December 1978 virtually alI Iranian oil exports had halted. However, In terms of total free- world of I product ron, the Iran! an Revol utlon did not result In u [ [ c [ [ [ c r b [ [ 0 c 6 G r u c ~ [ 1 0 large~ productJon losses due to production Increases In non- disrupted countrJes. While total free-world production was reduced by about 4.5 p~rcent between December 1978 and January 1979. free- world production was very nearly back to Its pre-disruption level by March 1979. Therefore, as was the case In 1973-74, the disruption was not long term or severe. However, the oil price changes that resulted during this time period were severe. Between July 1978 and January 1981 the official price of Saudi I ight crude increased from $12.70 to $32.00 per barrel. Various arguments have been put forth to explain these huge escalations. Certainly, production reductions by Saudi Arabia during this time period contributed to the increases. Further. same have argued that between the,dJsruption of 1973-74 and the Iranian Revolution of 1978 there was a major structural change In the role of the major oil companies in the production and distr!but!on of o!l from the OPEC countr!es. It !s argued that less control over the d!str!but!on of available ol I by the major o!l companies led to !ncreased competit!on for available oil on the spot market. drastlcal ly Increasing spot pr!ces, and eventually leading to official pr!ce escalations. Whatever the underly!ng reasons for the pr!ce !ncreases, same obvious market structural changes occurred during the 1973 to 1981 ' time period. Ff rst. OPEC, and the OAPEC countr!es !n part!cular. ga!ned tremendous market power as compared to the prev!ous era. No longer was OPEC an organization that attempted to siphon off o!J wealth that had previously gone to consum!ng countr!es and major or 1 compan res. OPEC, or at least a subset thereof, was now the r u [ [ r [ [ [ c r 6 [ [ 0 c c u r l c b [ 11 leader of the world ofl market wtth power to make productfon decfstons, Jf not Jn fact to set oil prices. Second, the OPEC countries wtth thefr newfound otl wealth had begun ambitious development programs that not only changed the nature of their economic and social structures, but also demanded the continuation of large Infusions of oil revenues If those programs were to continue. Third, the major oil companies had lost much of their abl I tty to control the dtstributton of crude and in so dotng had lost some abll tty to "smooth out" short-term and relatively minor supply disruptions. Ftnal ly, oil consumers and non-OPEC producers had begun the long adjustment process to the huge price Increases of the 1970s and early 1980s. However, the oil producing nations clearly maintained control of the world oil market, despite the facts that oil consumption was betng reduced through capital replacement, factor substitution, etc., and non-OPEC production was on the rise. The rel-ative power of the players In the determination of world oil prices had changed, and with that change came drastic escalations In world oil prIces. C. The 1981 to Present Time Period Although throughout the 1970s the price of crude on the world market appeared to be downwardly rigid --at least In nominal terms --a series of events began around the first part of 1981 that suggested a reduction In the strength of OPEC In relation to non- OPEC o II producers and major consumIng countries. In September 1980 t lghtl ng began between two OPEC members -"':' I ran and Iraq -- resulting In drastic reductions In exports from those countries. c [ [ r [ [ c c r 6 [ [ c r E c c t r b [ 12 lrant~n exports decreased by about 1 mil lion barrels per day-- Iraq's by about 3 mil !Jon barrels per day. While the size of the disruption In . percentage terms was about the same as that accompanying the Iranian Revolution --about 4 to 5 percent of free-world production --the Impact on the price of oil was significantly less. There are several reasons prices did not escalate as during previous disruptions. Certainly, the record level of oil stocks at the outbreak of the war contributed to the calm. Further, the consumption of oil In the world market was decreasing as a result of a worldwide economic slowdown and as a result of the addition of more efficient fuel use capactty. For example, U.S. consumption of refined products decreased from a peak of 18.8 mil lion barrels per day In 1978 to 15.3 mJI lion barrels per day In. 1982. World consumption decreased from a peak of 65.1 mil I Jon barrels per day In 1979 to 58.9 mil lion barrels per day In 1982. In addition, non-OPEC producers had responded to the huge price Increases of the 1970s by sharply Increasing their own production levels. While total OPEC production decreased from a high of about 31 million barrels per day In 1979 to 17.3 mill ton barrels per day In May 1984. production In non-OPEC, free-world countries Increased from 17.6 to 21.5 mil I ton barrels per day during the same time period. In addition, the Soviet Union Increased oil exports to the free world from 1.28 to 1.54 mil I ton barrels per day between 1980 and 1982. The pressures exerted on OPEC's pr I cl ng structure by non-OPEC production Increases and decreases In worldwide oil consumption resulted In an Increasing discrepancy between official OPEC prices C [ [ ~ [ r [ (, , -~ .; r 6 c c c c [ u r L [ G [ 13 and crude prf"ces on the spot market. In response to these pressures OPEC met In March 1983 and announced a $5 reduction In their benchma~k price from $34 to $29 per barrel. At the same meeting OPEC agreed to set an overal I output cell lng for the organization at 17.5 mil lion barrels per day, with the majority of the production constraints being borne by the major oil producers of the Persian Gulf. While the actions taken by OPEC at the March 1983 meeting helped to stabll lze oll prices and prevent further erosion of OPEC's powerbase, other problems threatened OPEC's future control of the world oil market. For example, to maintain the ambitious development programs begun by many of the oil producers In the 1970s, large Inflows of oil revenues were needed. However, because of Increased competition from non-OPEC producers, oil demand reductions, and the official OPEC price reduction, same development programs were threatened. These pressures resulted In Increased compet!t!on among the OPEC members for oil sales. In recent years, pressures have been most severe In the OPEC countries w!th larger populat!ons --e.g., Iran, Iraq, lndones!a, Nigeria, and Venezuela. In addition, both Iran and Iraq have .required large Inflows of oil revenues to maintain their continuing war. Accord!ng to a recent report <Chase Manhattan Bank, 1984) OPEC as a whole experienced an account def lcf t of $16.5 bIll !on In 1982 and $21.5 b i Ilion ! n 1983. As a result of these and other pressures, OPEC has In recent months experienced difficulty In maintaining Its offTclal production ceiling of 17.5 million barrels per day. In the last quarter of 1 983 total OPEC product ron exceeded 19 mil I !on barrels c c [" r~ b [ [ r -, ) c r l-.r [ [ G c c [ f' L c 6 L 14 per day putting additional pressure on OPEC's official price· structure. More recently, certain OPEC members have officially or In effect reduqed their ·prices below the official price levels Ntgerla, Iran, and Iraq betng cases In point. Pressures on official OPEC prices from overproduction continue at the time of this writing. For example, price reductions by Norway and Great Britain have placed additional pressures on Nigeria which produces crude of a similar type. Therefore, since 1981 there have been changes In the structure of the world oil market that have Impacted. world oil prices. Reactions by non-OPEC producers and oil demanders to the severe price Increases of the 1970s forced both reductions In official OPEC prices and the Imposition of production cell lngs on OPEC members. As a result, oil revenue reductions have threatened development programs and In some cases government stab! I tty --for example, the recent coup In Nigeria. Further, the threat of severe and sudden oil price escalations resulting from supply disruptions has diminished as productive capacity far exceeds production. Although to date the core members of OPEC--mainly Saud! Arabia-- have maintained their base price of $29 per barrel by absorbing the majority of the necessary production reductions, several OPEC members continue to cheat In terms of both prices and production I eve Is. The cohesion of OPEC In the post-1981 era has been shaken c [ r b c L c fl c ~ b [ c D [j 6 lJ r u [ c u L 1 5 and w1th that change have come sfgnfftcant reductfons fn world otl 2 pr fees. Table 2.1 gtves the h·Tstorfcal changes fn the offfcf al prJce of Saudt Arabfan I Jght crude and Ntgerfan crude from 1973 to the current tfme perJod. The prfces of other crudes vary dependfng on thetr respective qual ftfes. II I. Methodologies Developed to Explafn and Forecast OJI PrJces The methodologies that have been developed to represent the functtonfng of the world ott market draw sTgnTffcantly from the modeler's dfsclpl tne and hfs or her perceptfon of htstory. At the rfsk of oversfmpl Tftcatton, thls · sectton summarfzes three matn categories of models that have been used to explatn hfstorfc world or I prfce changes .and, In some cases, used to forecast future ott prfce trends. As stated earl fer, an Tndepth account of each model will not be attempted because of the sheer number of models and because such an exercfse· would probab I y, for our purposes, contribute I Tttle to a better understanding of world oil price forecasts. Rather, models or methodologfes are placed In one of the three major groups dlscussed above--(1) economic models, (2) pol !tical models, and (3) economlc-pol Ttlcal combination models. FTgure 2.1 gfves a graphical representatlon of the disaggregation ~Tstorfcal accounts of developments In the world oil market are available from numerous sources. See, for example, Deese and Nye (1981>, Vernon (1976), Landis and Klass (1980), Bohland Russell (1978), Curlee (1983>, Johany (1980), and Griffin and Teece (1982>. For more recent assessments of the world ol I market and how recent trends may signal future changes In the structure of the market, see, for example, Bohl and Quandt <1984), Horwich and Weimer (1984), Weyant (1983), Kash, Fox, and Wilbanks (1983), and Curlee ( 1984). c [ [ c [ [ [ [ r 6 [ [ c l [ [ r: L t L - l .. 16 Table 2.1. Official prices of Saudi Arabian light and Nigerian crude for selected dates · (Nomi~al dollars per barrel) SAUDI LIGHT NIGERIAN TIME JANUARY 1973 $2.10 $3.10 DECEP.eER 1973 $3.60 $5.84 JANUARY 1974 $9.60 $12.60 JULY 1974 $10.40 $11.85 JANUARY 1975 $10.46 $11.80 JUNE 1976 $11.51 $12.89 JUNE 1977 $12.70 $14.63 JULY 1978 $12.70 $13.87 JULY 1979 $18.00 $23.49 JULY 1980 $28.00 $37.02 JULY 1981 $32.00 $39.92 JANUARY 1982 $34.00 $36.52 MARCH 1983 $29.00 $30.00 JULY 1984 $29.00 $29.85 SOURCES: AmerJcan Petroleum Institute (1984> Central Intel! !gence Agency <1984) Cur I ee ( 1 983 > c c [ h c [ [ fJ ~ 6 [ c 0 c c u r: L J~ L E L .. ECONOMIC ----1 MOOELS OPT I M I ZA Tl ON 17 -f PROFIT MAXIMIZING COMPETITIVE REVENUE MAXIMIZ lNG 01HER CB JECT IVES -L ROFIT MAXIMIZING MONOPOLISTIC 01HER <BJECTIVES SI~LATION--{CAPACITY DRIVEN 01HER FO~AL POLl TICAL MODELS [INFORMAL [FO~Al ECONOMIc-POLITICAL -------1 .· COMBINATION MODELS INFORMAL Fig. 2.1 Methodologies to forecast world oil prices c [ [ l [ r [ c r L [ [ c r E 5 f' L [ c [ 18 4 of the three major model lng types Into various subcategories. The key market and non-market parameters that are hypothesized to Impact oil pr.lces under each group of methodologies are subsequently Identified. A. Economic Models Economic models of the world oil market can be divided Into two broad subgroups --optimization models and simulation models. In optimization models some part of the oil market Is assumed to have foresight of future market conditions and alters their production and/or pricing decisions to maximize some given obJective function. For example, In many models the objective function Is discounted profits. In simulation models price changes are assumed to be determined by some set of market or non-market parameters over which no expl !cit control Is exercised by any subset of the ol I market. For example, In many simulation models It Is assumed that world oil prices are determined by the level of excess production capacity In the OPEC countries or a subset thereof (where the level of production capacity Is set exogenously>. 1. Optimization Models The majority of the optimization models assume that OPEC, or a subset of OPEC, sets production In order to maximize the stream of discounted profits from the production of their oil reserves. The selection of the optimal production stream is dependent upon numerous parameters, the Identification of which requires a brief review of the economics of depletable resources. n c [ r: tJ C [ [ c r L [ c G ~ 6 [ r lJ f!. L L l The depletable resources semtnal article by Hotel ltng. 19 I tterature Is based on the 1931 In that article Hotel I lng describes how a resource _owner with zero production costs wll I produce over time under both competitive and monopol Istre market conditions. Fran elementary economics It can easily be shown that In a competitive market price wll I be equal to marginal cost. However. In the case of a depletable resource the marginal cost of the resource Is not just the marginal cost of production. In essence the owner of a depletable resource Is faced with the question of when to produce the resource. Therefore. In aqdltton to the actual cost of production. the resource owner must also consider the opoortyntty cost of producing the resource today at the expense of not producing the resource tomorrow or. alternatively. producing tomorrow at the expense of not producing today. In the slmpl lfled case that Hotel I lng considered where production costs are zero and resources are fixed. It can be shown that the price of the depletable resource must rise at the rate of Interest In a competitive market. Intuitively. this conclusion Is obvious because In equl I lbrlum the value of the resource In the ground must be equal to the value of the resource being produced. The owner of the resource has the option of producing the resource today and Investing the payment for that resource at the market rate of Interest. or the owner can leave the resource In the ground and have the value of the resource Increase at the market rate of Interest. In equl I lbr!um the resource owner wll I be Indifferent between the nvo chol ces. c [ [ p [ [ [ [ r: L [ c D c [ [j f' l c . ' L [ 20 In the case of a monopoly that controls the resource base, It can be shown that the monopol 1st <under the same slmpl lfylng assunptJons gJ~en above-) wrt.l maxlmlze profits when marginal revenue, which wll I be less than price, rises at the market rate of Interest. This Is a straightforward extension of the argument for the competitive case wfth the addftfon of the profit maximizing condition for monopol lsts --r.e., proffts are maxfmlzed when marginal revenue Is equal to marginal cost. An Interesting extension of the monopoly argument shows that. the monopoly factng a demand function with price elastfclty other than unity will produce less and price higher In the first time periods of production and wll I produce more and prfce lower In the later time periods of productfon as compared to a competitive market. The monopoly thus pranotes resource conservatl on. Since Hotel I lng's seminal artfcle, numerous papers have extended this baste methodology to account for some of the sfmpl lfylng assumptions made by Hotel I lng --for example, by Including positive and Increasing resource production costs, accounting for the posslbl I lty of reserve additions, or Including a backstop technology or good that would replace the resource In question at sane backstop price. Depending on which assumptions are relaxed, the analysis can become quite compl lcated. Again at the risk of overslmpl lflcatlon, the major parameters that determine the price of the depletable resource under profit maximizing conditions can be categorized In a few groups. The major parameters ( 2) the rate of of concern are (1) the size of the resource base, Interest or the discount rate, (3) the backstop c [ [ r [ [ [ c r L [ c c E t= l; 8 f' u c [J [ 21 price; (4) the long-run and short-run price elasticity of demand and supply of the resource, (5) the rate of world economic growth and how that g~owth translates Into changes In demand for the resource, and (6) the speed at which the market adjusts to changes In market prices and economic activity. In the mJd 1970s the economics profession general Jy modeled the world oil market by assuming that OPEC or the core countries within OPEC--I.e., countries such as Saudi Arabia that have vast and low-cost reserves as compared to other OPEC countries effectively cartel !zed the market. It was assumed that by restricting production these producing countries could In effect set the price of world oil such that their discounted profits were maximized. Supposedly, once the oil cartel gained control of the ol I market In the early 1970s, production and pricing decisions were made by the producing countries so as to move from the previous competitive price and production trajectories to the hlgher-prtced and lower-production trajectories dictated by monopol Istre profit maximization. Models that fal I Into this cat~gory Include Pindyck (1978), Kalymon (1975), Cremer and Weitzman (1976), Gately, Kyle, and Fisher (1977), and Singer (1982). There are also numerous models that are variations of the baste monopol lstJc profit maximizing approaches. Hnyll lcza and Plndyck (1976), Eckbo (1976), and the Salant/ICF Model (see EMF, 1982) address the Inherent problems faced by any cartel that arise from bargaining over pr!c!ng and production decisions. Because of varying economic and social concerns In different countries within the cartel, different members wfl I have different perceptions about c c [ b [ C' t_~ [ t r 6 [ [ u c c c r L [ b [ 22 which production· and price paths are optimal. These models use game theory to study bargaining within the cartel and Identify side payments that. ~!¥ be necessary to maintain the cohesion of the cartel. In contradiction to the methodologies discussed above, numerous modelers have argued that one does not necessarily have to assume that OPEC or a subset of OPEC acts as a monopol lsttc profit maximizer to explain historic oil price Increases or to forecast future price trends. MacAvoy (1982) has argued that the prtce Increases of 1973-74 were unavoidable even under competitive conditions, given the very tight oil markets that had resulted from very low historic oil prices. Some modelers argue that the drastic historical price changes can be explained within a competitive market In whtch the control of the resource base shifted from International oil companies to the oil producing countries. For example, Johany (1980) has argued that the price Increases of the early 1970s were a natural result of the transfer of the property rights to ol I from the multinational oil companies to the producing countries. It can be argued that the International oil companies made decisions based on a high dlseount rate previous to the early 1970s because of fears of expropriation. Once property rights were transferred to the producing countries, their lower discount rates naturally translated Into lower levels of production and higher ol I prices. Another variant of the competitive argument Is based on the abl I lty of the producing countries to absorb the revenues from their oil sales. This Is the basic approach taken In Teese (1982>, c c [ c C [ [ 0 r L [ c 0 8 E c r u u c [ 23 Cremer and SalehJ-Isfahanr (1980>, Mead (1979), and EzzatJ (1976). These models Jn effect argue for a backward bendtng supply curve for the major oJI producing countries. The htgher the oil prtce goes, the less the major producing countries produce because they have no abtl tty to uti I Jze the Increased revenues. This position has been criticized on the basts that the producing countries must assume that oil In the ground ts worth more that "money Jn the bank." The proponents of this positron counter this crlttcrsm by arguing that Investments made by producing countries In consumtng countries are always subject to threats of .expropriation, thus decreasing the relative value of "money In the bank." Other variations of the competitive framework also exist. Bl ltzer, Meeraus, and Stoutzesdyk (1975) argue that the major producing countries attempt to satisfy the dual objectives of the maintenance of market share and high current revenues. In the ETA/Macro Model (see EMF-6, 1982>, the basic objective Is to maximize the discounted uti I tty of oil consumption In OECD <Organization of Economic Cooperation and Development) countries. 011 prices are a function of OECD Imports. In the Kennedy-Nehrlng Model (see EMF-6, 1982) It Is assumed that the competitive non-OPEC conventional oil producers have maximize discounted profits. exogenously. 2. Simulation Models perfect foresight and act to OPEC production Is determined The majority ot models used today to forecast world oil prices do not employ the optimization approach discussed above. Rather, simulation models --wnfch do not dfrectly assume that a subset of c [ [ c [ [ [ c r L; [ [ c E E c r u [ L [ 24 the ol1 market manipulates oil production or prices to satisfy some obJective--are used to forecast oJI prices and other key oil market paraneter_s. These slmul atlon models often empl C¥ a simp I lstlc methodology In determlnfng world ofl prices--as compared to the optfmlzatfon approaches discussed price determination above. The most common assumption about oil Is that prices rise when capacity uti I lzatlon In the OPEC countries rfses above some prescribed level --usually about 85 percent. At some capacity uti I lzatlon level below 85 percent the price of oil Is assumed to decl lne. The level of OPEC production capacity fs usually an exogenous Input. Models that fal I within this category fnclude Gately (1983>, Levy (1974), OMS <DOE, 1983), OILMAR <see EMF-6, 1982), DRI (1983>~ WOIL <see EMF-6, 1982), Braden <198U, Gately, Kyle, and Fisher (see EW-6, 1982), OpeconomTcs <see EMF-6. 1982), and OTitank <see EMF-6, 1982). There are also models that have varlatfons of this baste theme. For example, EMF-6, 1 982) oil companl es the International Petroleum Exchange <IEP) Model (see expl fcftly considers the actfons of the multfnattonal In Its slmulatton. Prices react to changes In the reserves to productton ratto, production ~osts, and exogenous royalttes. The IEP fs the only simulation model reviewed In which the production capacity of OPEC Is not set exogenously. The key parameters that determine world oil prices are usually the same In simulation models as In optimization models--I.e., oil reserves, Interest rates, backstop price, economic growth, adjustment rates, and short-run and I ong-run sup pi y and demand· elastlcltfes. Econometric approaches are used to Identify how c [ [ r [ [ [ c r L [ [ c E E c r~ L [ [ L 25 these parameters --tn addttton to OPEC capacity uttl tzatTon -- have been rel~ted to htstorlcal ott prtce changes. Assumptions about how thes~ key parameters mav change In future years allows the modeler to project future otl prtce trends. Simulation model·s can be crtttclzed on the grounds that the major determinant of oil prtces --OPEC capacity uti I lzatlon --Is usu~lly assumed to be exogenous.3 Because of thts "overslmpl tftcatton;" It can be argued that most slmulatton models are more adept at forecasting how oil market parameters other than prtce m~ react to future prtce changes, rather than actually forecasting world otl prices. B. Pol lttcal Models Some modelers have made the lmpl felt assumption that pol tttcal factors dominate economic criteria In determining the price of oil In the world market. To some extent It can be assumed that pol ltlcal goals within the major producing countries are compatible with wealth maximizing goals. However, It has been argued that often economic goals--e.g., wealth or revenue maximization-- must be, and are, sacr!ftced to obtain both Internal and external pol lt!cal objectives. In these cases, It can be argued that formal or Informal pol lttcal models can best suggest why prices have changed htstortcally and how ell prices may change In the future. Two pol !tical models are reviewed briefly here. In Moran (1982) it Is argued that the pol !tical decisions of Saudi Arabia-- ~urlee (forthcoming) argues that the qual lty of the existing data on current and projected OPEC and non-OPEC production capacity Is poor, thus placing empirical projections from these models In question. c [ [ c C [ [ [J n L [ c 0 E r c r: b F u [i L [ 26 the assumed OPEC leader --wll I have a significant Influence on future world oJI prices. This Informal model argues that Saudi ArabJa attempts_to maximize Its own pol ltlcal objectives In Its oil production decisions, while being constrained by potential hostile pressures both Internally and externally. Moran presents a detailed description of the probable pol ltlcal objectives of the numerous pol ltJcal groups within Saudi Arabia that may have Influence over future energy pol Icy decisions. Moran thus presents a model of how differing Internal pol ltlcal objectives--Including wealth maximization on the part of some power centers Tn Saudi Arabia--can Interact to determine oil production dectsJons that do not conform to any of the criteria specified In the economic models discussed above. In a more formal model by Saaty and Gholamnezhad (1981) a methodology called "analytic hierarchies" Is used to formulate a pol Jtlcal model of the world oil market that takes a more global perspective. 4 Several pol !tical factors are considered, Including the degree of lnstabll lty In the Persian Gulf Region, the Intensity of the Arab-Israel I conflict, and the Increased Influence of the Soviet Union In the Middle East. Several "econanlc-technologlcal" factors are also considered within the pol !tical framework, Including strategies of the consuming countries to Influence oil consumption, excess-or !-production capacity, the Influence of the International financial Institutions, oil discovery rates, and development of alternative energy sources. 4 The "analytic hierarchies" approach Is not explained In great detail by the authors. For a detailed description of that approach see Saaty ( 1 980) • t [ r L 0 [ [ [ 6 r L; [ c c c E c f lJ B L L 27 .. C. Combination Models In recent years there has been a growing perception among many modelers of the world oil market that the mechanism by which world oil prices are determined Is far more campi lcated than that represented In either economic or pol ltlcal models. This real lzatlon has .followed the poor predictive abll ltles of both economic and pol ltlcal model Jng approaches. The frustration felt by modelers Is best explained by two quotes from one of the foremost researchers In the field, Robert Ptndyck. Plndyck was one of the first authors to model OPEC as a monopol Istre wealth maximizer. Plndyck's confidence In this approach Is reflected In his 1979 paper. "So far OPEC has done more or less the same thing, and to the extent that It continues to do so, economic maximization should provide a reasonable basis for explaining and forecasting the prf ce of oil" (page 175). However, In a pub I I cat I on by PI ndyck ln 1~82 he reassesses the worth and predictive capabll !ties of "models that describe 0 economlcally rational' price formation by the OPEC cartel" (page 175). "I wIll argue that Improved models might be useful for examlnlg various theoretical and empirical Issues In the behavior of ol I markets, but that they are not needed, and would not be very useful, for predicting world oil prices" <page 176). "It seems to me that from a theoretical point of view, models of OPEC oil pricing have reached a practical I lmlt as tools of analysis. As far as empirical predictions of oil prices are concerned, some of these models have already exceeded that I l~lt" <page 179>. c [ [ c [ [ [ c r L [ c c c E c f u b [ l 28 MOdel lng methodologies have therefore evolved to Incorporate the real Izatt on that future ol I prIces are very uncertal n and difficult to p~edtct. On the one hand, some modelers have adopted stmpl lstlc "rules of thumb" to forecast or I prtces --such as the assumption used In most stmulatlon models that otl prices are a function of the degree to which OPEC production capacity Is uti I lzed. On the other hand, some modelers have (1) attempted to Incorporate pol ltlcal factors Into their economic approaches or <2> have opted for Informal methodologies that lmpl tcttly Include economic and pol ttlcal criteria. An example of the first type Is a model by Daly, Griffin, and Steel (1982). In that model OPEC supply responses are modeled on the basts of actual and potential reserves, absorptive capacity, and pol lttcal constraints. The model lng approach --which employs a baste economic approach constrained by perceived pol !tical real ltles --was used to assess the stab! I lty of the OPEC cartel given different price paths. Examples of the second type are Plndyck (1982) and Weyant (1983). In these papers no formal models are constructed. Rather, projections of future market conditions are based on Intuitive judgments based on the authors' perceptions of economic and pol Jtlcal real Jttes. I V. Co nc I us l on s Several conclusions can be drawn from this review of methodologies designed to explain and forecast world all price changes. Fl rst, the roots of the various approaches are In the modelers' dlsclpl lnary perspectives and perceptions of hlstortcal c [ [ l [ L [ c r L_; [ [ c c t c r~ L [ c l 29 even~ In the world oil market. As has been reviewed, the world oil market has gone through many structural changes In the past two decades. Oepe~dlng on the time period that Is being modeled, the particular methodology that Is required mav differ. Second, the methodologies have evolved during the past decade. This evolution has come as a result of the poor predictive capabl I ltles of previous models and because there has been a real lzatton that the structure of the oil market--Including the relevant players, the obJective functions of the players, and the abll ttles of the players to manipulate the market to maximize those obJectives-- has changed. Finally, and most Importantly for our purposes, there Is no consensus about what methodology should be used to forecast future world oil prices. To some extent there has been a movement toward a consensus that prices wll I be determined by both economic and pol ltlcal factors. Beyond that there Is Jess than complete agreement about the particular model lng approach that should be used. However, It Is generally agreed that different model lng methodologies may 5 forecasts. lead to significantly different oll price Whatever the conceptual problems that remain Jn the area of oil price forecasting, there Is no shortage of price forecasts from which to choose and evaluate. A review of the recent oll price forecasts and their underlying assumptions Is the subject of the next chapter. 5 For more Information on the different methodologies that have been developed to foreCast world or I prices see, for example, Sweeney <1983), Bercer <1981 l, Griffin and Teese <1982), and Gately (1984). J J l n j J J j D J J J 0 J J J ~ J J j c [ [ l [ [ [ 0 r L_, [ [ c c E c F ld r: L: c [ 31 • 3. WORLD OIL PRICE FORECASTS I • In tro<luct I on The purpose of this chapter Is to review and summarize recent mid-to long-term world oil price forecasts. As was the case In the previous chapter. the large number of oil market models and forecasts prevents a detailed discussion of each forecast. Rather, the various forecasts are summarized and. where possible, categorized according to the model lng types and major model parameters discussed In the preceding chapter. Of particular Interest are the variations In the forecasts caused by changes In the key parameters. In addition, the forecasts are discussed In terms of the dates the forecasts were made. A key question Js whether the most recent oil price forecasts those completed after the official oil price reduction of 1983 dIffer slgnlflcantly from earlier forecasts. Because of the l_arge number of forecasts available. only those completed after 1980 are considered In this review. A direct comparison of different price forecasts Is difficult for several reasons. In many cases only the forecast Is given. The assumptions about key model parameters. and In some cases the baste model lng methodology, are often not given. Further. because the model tng methodologies and assumptions about key parameters differ so widely In some cases, It Is Impossible to pinpoint why price forecasts differ. Because of these problems the available forecasts are grouped Into three categorres. The first contains what m~ be termed "expert judgment" in that etten no rigorous mathematical model Is [J [ [ t [ [ ~ c r: ~ [ [ c E b c r~ u l c l 32 used fo make the forecasts, or If a formal model Is used the forecasT Itself has been "flavored" with the forecaster's own subjective view~ of the market situation. Often these forecasts do not give specific price projections, but rather project general prtce trends. This class of forecasts contains predictions based on the Informal economic-pol ltlcal methodologies described In the preceding chapter, as wei I as Informal economic and pol ltlcal models. The second category contains forecasTs thaT are derived from formal mathematical models. Many forecasts In th r s category give the basic methodology and values of the major InpuT parameters used to obtain the forecasts. In some cases the forecasts are tested for their sensitivities to changes In some of the key parameters. The third category contains surveys of nunerous Individuals and groups that periodically make price forecasts. These forecasts are based on formal mathematical models, expert judgment, and/or combinations of the two approaches. However, the surveys do not give sufficient Information about the Inputs to the Individual forecasts to make detailed comparisons of the Individual forecasts. I I. Forecasts Based on Expert Judgment As Is obvious from the preceding chapter, there are a variety of approaches to model lng the world oil market. Likewise, there are numerous positions about future market trends. In this section the discussion of those positions and forecasts begins with a review of several quotes from lndlvl.duals well noted In the field: " ••• ( S) o I ong of production direction, and as the OPEC nations maintain the current system control, the system ls unstable In the upward a price hlke Is almost guaranteed at any time [ [ [ E [ [ [ ~ r '---.) [ [ c E b c p L;<. E [ [ 33 u~less the core nations take active roles toward preventing lt. For these reasons, I would expect stll I higher (real> prices In the 1980's despite weak demand" <Adelman, 1982. page 54). "OPEC will continue to have power over price, especial Jy In the short term,· and Its power wll I Increase when Its capacity uti I lzatlon Increases. But, over the longer term, taking ten- year or twen1y-year averages, OPEC's market power w II I be constrained by the underlying price-responsiveness of demand and of non-OPEC supply, for oil and alternative energy sources" <Gately, 1984. page 1113). "Considering that OPEC has yet to demonstrate that It has the wherewithal to del lmlt competitive output expansion, two decades of constant real prices Is a strong possibility--In the absence of a supply disruption of significance" <Teece, 1982. page 86). "In the long run, the desire of the Saudis to avoid the vul nerabll lty that the mere existence of huge export capabl llty brings--vulnerability to consumer government pressures to use It, vulnerabll lty to producer government pressures to Jet It I le Idle --exercises a great dampening effect on capacity plans for the kingdom, Irrespective of what huge reserves and low discount rate might Indicate •••• <T>he analysis of Saudi decision-making presented here suggests a more pessimistic view of oil prices <pessimistic from the point of view of energy consumers> than the economic optimization approach for any given set of assumptions about supply and demand for energy" <Moran, 1982. page 116). "It may be assiiTied •.• that OPEC's pricing policy In the 1980s wll I be more sensitive to market conditions than It was In the 1970s. If this Is so. OPEC real prices will not rise significantly over the next nine years. If this assumption Is wrong oil prices could continue to rise significantly for only a I lmlted period before the organizations's price structure wou I d col I apse under the ons I aught of consumer conservatl on and producer competition" <Uchtblau, 1982. page 143). "While there remains a high degree of uncertainty about future world oil prices, most of the uncertainty concerns not whether real prices will rise, but rather how rapidly they will Increase. • •• The overal I real price trend wit I be upward" (James Sweeney as quoted In Daly, Griffin, and Steele, 1982. page 145). "Eventual ly, the upward pressure on oil demand caused by Increases In econanlc activity could more than offset the downward pressure· resulting from the dwindling adjustment to the 1979-1980 price Increase. . •• The slack world of I market me/t! persist tor t-Iro or three more years, but Is unlikely to I ast much I onger" (Weyant, 1 984. pages 393-394). " [, [ [ l' [ [ [ ~~ ti r l_c [ c 0 E E G r l_j [ [ [ 34 "1'he results of this paper suggest that the Iranian pol ftfcal upheavals of 1 fJ78-79 and the subsequent do ubi I ng of crude of I prfces mav wei I have defined the I lmlts of OPEC's monopoly power. A long run real price path significantly greater than $32 per barrel seems I lkely to evoke large suppl les of synthetic fuels, coupled with substantial conservation effects --events which, taken together, make such a price path unl lkely" <Daly, Griffin, and Steele, 1982. page 173). "<l>t would probably be reasonable to assign a 2 to 4 percent real rate of growth to future oil prices as part of a 0 most I lkely' forecast. But at the same time It must be remembered that the confidence Interval around that forecast Is extremely wide, perhaps as large as 50 to 100 percent. What should really matter In terms of the decision of energy producers and energy consumers Is not the 0 best guess' forecast, but the fact that the uncertainty around that forecast Is considerable" <PI ndyck, 1982. page 179). "<T>he authors In this volume bel leva that the world economy has by no means fully responded to the 1973-74 price shock, let alone the 1979 shock" <Griffin and Teece, 1982, page 208>. "<T>he target revenue model suggests that substantial downward price movements are I fkely In the event of a prolonged soft market" <Griffin and Teece, 1982. page 211). "The consensus of the authors appears to be that If a large price hike occurs In response to some pol ltfcal disruption, It wll I not be sustainable In the long run, unless such pol ftlcal upheavals permanently take appreciable capacity out of production" <Grlffln and Teece, 1982. page 212). "At most, after 1990 I would expect a real Increase of percent per year ln oil prices after the $34 price Is attained" <Netschert, 1983, page 141 >. "If OPEC operates as a revenue-max! ml zing cartel, It w !II I ower nominal oll prlces to the $25--29/bbl range and allow Inflation to reduce real oll prl ces l n 1986 to the $20-23/bbl range" ( Roumasset, Isaak, and Fesh arak I, 1 983, page 1 93 > • "Unllkethe sltuatlonln the 1970s, when the Interest of exporting and Importing countries seemed to be diametrically opposed, lt now appears that OPEC and the Importing countries will both lose If prlces rlse. • •• Consequently, It ls reasonable to Imagine that most OPEC members wll I recognize thelr sel f-lnterest ln stable prlces In coming years" <Bohl and Quandt, 1984, page 18). "If these models and scenarios correctly futures, we can expect the real price of rernal n stagnant tor several years and significantly ... " <EMF-6, 1982. page 92). represent world of I ol I to decl lne or then to rise c [ l~ t [ [ [ 8 . . p Lc~ [ c 0 E n t; c F y [ [ [ 35 4 "Whether fran operations wlthfn OPEC or those In an open market, world crude otl prtce should remain constant or increase at most by one or two percentage points each year during the t980s" <MacAvoy, 1982. page 78). "The OPEC countries ••• --Algerta, Indonesia, Nigeria, Venezuela, and Ecuador --wll I advocate large oil price Increases during the 1980s because of their need for oil revenues and because of their I lmlted role as exporters In the long term. • •• Saudi Arabia and UAE have an economic Incentive to block Increases--which would bring the oil price above ••• the price at which substitutes to OPEC oil wll I start being Introduced ••• or the price at which the value of OPEC's ftnanclal surpluses wll I begin to deprectate ••• " <Aperjls, 1982. page 125>. While the revtew of these forecasts by tndustry experts does not lead to any deflntte conclusions, It does suggest a movement toward a consensus on several potnts. First, the predominant view Is that real oil prices wll I not Increase or decrease significantly durrng the remaInder of the t 980s. In the 1990s and beyond real prices are expected to Increase moderately, but should not be Subject to the drastic price Increases observed In the 1 970s. Drastic price jumps are not seen to be In the Interest of either oil const.mers or OPEC. Second, the prevail tng vtew Is that OPEC wll I continue to maintain a significant degree of control over the oil market, but to a lesser restraint on OPEC's power adjustments of oil consumers degree than rn the 1970s. This will result from the continuing and non-OPEC oil producers to the drasttc price escalations of the 1970s. Thlrd, there Is a general consensus that the Impacts of supply disruption, such as those of the 1970s, may result In short-term price jumps, but those jumps would probably not be maintainable over the long run. Finally, ·while price forecasts do not diverge as widely as they have In the E [ [ e [ r [ u r: L,, ("""" _, c c [ h [.~ ' ,1 p w L: L l 36 past,• there Is stll I a recognition and a warning that world oil price forecasts contain a great deal of uncertainty. Ill. Forecasts Based on Formal Models In this section numerous price forecasts that have resulted from formal mathematical models are reviewed. The sources of these forecasts vary widely --from academic Institutions, to government agencies, to private firms. In addition, the forecasts vary Tn terms of the year they were publIshed and In terms of the model lng methodology used. At I east one sample of each of the formal methodologies discussed In the previous chapter Is represented. In some cases only one observatTon --I.e., one forecast --was available. In other cases there are as many as four observattons from a particular model. The review of forecasts presented here certainly does not represent alI available forecasts. The review does, however, give a sampl lng of different model Tng types, forecast years, and forecasting Institutions. Further, this review does not lead to definitive statements about how and why forecasts vary due to differences In underlying assumptions and parameter specifications. Some broader general conclusions can, hOttever, be drawn. In the following subsection the forecasts are presented and discussed In terms of their general model lng methodology, date of publ lcatlon, and, where possible, the specific model used. That review Is followed by a discussion of how changes In some of the key model lng param.eters have been shown to Impact some of the forecasts. f ' ,J c [ fj [ [ [ e r u [ c 0 D [ c f fL.\. 6 L [ 37 A. T~ Forecasts To help tn our assessment of the forecasts, the price proJections are arranged. accordtng to dtfferent parameters In different tables. Table 3.1 presents the forecasts arranged by the year the forecast was publ Jshed. Whtle the prtce projecttons are gtven Jn terms of constant dol Jars for each lndtvldual forecast-- as gtven Jn the referenced source--the dtfferent forecasts have not been adJusted to reflect a base-year constant dol Iars In this table. In Table 3.2 the forecasts are agatn arranged by year publ lshed, but alI prtce projections have been converted to constant 1981 dol lars.6 Forecasts are arranged by model Jng type and by the organization maktng the forecasts Jn Tables 3.3 and 3.4, respecttvel y. In Table 3.5 the forecasts are arranged by thefr projected 1990 price level. An overview Of al 1 the forecasts shows that oil prices--In constant 1981 dol Jars--are projected to be between $17.12 and $50.73 per barrel In 1985 Cwlth a medl an of $25.46), between $19.56 and $64.00 per barrel In 1990 <with a median of $36.65), between $24.57 and $59.62 dollars per barrel In 1995 (with a median of 6 AI I prices have been adjusted by the U.S. GNP deflator. While the use of a U.S. price deflator Is satisfactory to adjust oil prices to a base year from a U.S. perspective, It m~ not be appropriate Jn adjusting world ol I prices to a base year. Because of significant changes In the exchange rates among different currencies, It Is not clear that adjusting prices by a U.S. defl~tor wfl I give accurate real price changes for the world market. For example, Huntington (1984) has argued that while the real price of ol I went down In the U.S. between 1981 and 1982. the real price went up In several other OECD countries--e.g., Italy, Japan, France, and the United Kingdom --due to exchange rate varl at! ons. F 38 [ Table 3.1. A sumary of oil price forecasts [ arnnged by year Jli> 1 i shed .. ear t Organization SClln:e Date of r-txtel Constant ar tobdel of Mxi!l SCllrce TyPe 1335 19!ll 1995 am 3)10 am Dollars r tFE OOE (FE ].gl) NA 43 198) !XE~Aa OOE,()PPA 1981 ~ 46 55 74 1982 [" MITRE a MrmE 1981 ~ 76 1982 Sa~ & Ghll. Sa~ & Gtol. 1~1 Political 41.06 44.86 [ Bali Boh1 15S2 Caroi nation al.2 1972 ETA ~ro EW~ 1982 c~ Opt g).6 00.2 64.1 68.9 ~l r ··' Kemec:tf-Nehri ng EW~ 1982 c~ Opt ~.8 n.5 71.4 71.4 ~1 r: Salant ICF EW~ 1982 Mom Opt 55.5 71.3 ~.8 l.C6.l a31 {___; \01 Marshall a ·1982 Mom Opt 31.53 35.91 ~.85 42.17 56.66 1982 [ Fosterl Foster ~2 NA 34 ~ 49 1982 9-CAa ~ ~2 NA 53.8 66.5 103.4 1982 L Gately EM='~ 1~2 Sirrulation 52.9 71.7 71.3 93.8 l.Sl31 c IPE EM='-6 1~2 Sirrulation 37.2 54.6 l.Sl31 EM="-6 1982 Sirrulation 64 ffi.8 12).2 127.3 l.Sl31 OilJ.M c OILTPH< EM="-6 1982 Sirrulation 63 92.1 129.7 152.3 l.Sl31 ()1S EM="-6 1982 Sirrulation 46.1 l.Sl31 E Opeconantcs EW~ 1982 Sirrulation ~.7 41.5 1981 c \tOIL EM='~. ~2 Sirrulation 47.8 69.6 81.8 69.1 l.Sl31 Singerb Singer ~Mom Opt 17.8 21.9 Zl.4 32.2 1.98) f 9-[AC ·~ ~NA Zfi.3 'l/.09 32.34 37.5 :D.39 64.48 1.983 L Texacod Texaco I.Se3 NA Z1 (B 29 ]) 1982 li OeYarajan et al Devar. et al l9i3 Si111Jlat1Cil 32.8 b llU 101 [)U ~ Si111JlatiCil ]),09 li.1 44.12 51 l.Sl31 [ c [ [ c [ L [ B r L; [ c c E [ c f bJ t f' 6 [ 39 4 Table 3.1. (continued) Year Organizatioo sa..rce ~te of r.tldel Constant a-r.tldel of tb21 sa..rce TtE! ~ 1.9'}) 1995 2lX) 2)10 aJaJ Dollars CMS Hig, EIA 1!83 SintJlation 34 48 ~ CMi Lew EIA lSlO SintJlation 21 28 1!E2 CMi Mid EIA lSJD SintJlation 25 '5I ~ RtUI8Sset et al Rans et al 1~ sintJlation 17.12 19.56 25.47 1$1 ~c 9£.A ~NA ai.J 'l/.9 32.5 40 Ei) 8) 1$3 ~rd llU ~ SintJlation 'll.S 25 28 1$4 CMi Hig, EIA 1984 SintJlation ]).53 45.64 65.89 1$3' CMi Hi g, Draft EIA ~ SintJlation ]) 40 55 l984 CMi Lew EIA 1984 SintJlation 22.44 29.16 36.54 1$3 CMS Lew Draft EIA l984 S1ntJlation a4 25 ]) l984 ()f) Mid EIA ~ SintJlation ;!;.52 36.65 !il.49 1983 CMi Mid Draft EIA 1984 SintJlation 'l/.92 l) 40 1984 ~ reported in Foster, Burton and Haf1]eter (1~). bsinger Jllints art ttlat these are rot oil price forecasts, rut rather reflect tile prices that 'fOJld t:e cha~ I?( a profit JTBXinrizing oil cartel. CAs reported in Alaska Po.rEr ~rizy (1~). drnterpreted fran a graphical ~tion. [J [ 4 40 [ Table 3.2. A sunnary of oil price forecasts arranged ~ -· year ~1 ished (In coostant 1~1 oollars) 0 -Year Organization Soorce Date of r.tx!el Constant or~l or Mn!l Soorce Type 1~ 19ClJ 1995 ;m) a:no (!)20 Do 11 a rs [ OPE OOE CPE lE8l ~ 47.02 1981 [ llE!UPPAa OOE;1JPPA 1981 ~ 43.37 51.86 69.77 1981 MITRE a MITRE 1981 ~ 71.66 1981 [ Satzy & Ghol. Satt;y & Ghol. 1~1 Political 41.()) 44.86 1981 Bohi !3001 1$2 Cami nation 39.41 1981 B ETA Macro tW~ 1982 CCI'fl) Opt 5J.6 00.2 64.1 68.9 1981 c Kennec:tf-Nehri ng EW~ 1~2 CClll' Opt 56.8 77.5 71.4 71.4 1981 LJ Salant ICF . EM='~ 1~2 Mooo Opt 55.5 71.3 a3.8 1()).1 1981 [ ~ r.Brshalla 1982 Mono 0 pt 29.73 33.86 36.63 39.76 53.43 1981 Fostera Foster 1982 ~ 32.()) 35.83 46.2 1981 b SI-CA a 9-I:A 1982 ~ :D.73 62.7 97.5 1981 c Gately EM='~ 1982 Sirrulation 52.9 71.7 71.3 93.8 1981 IPE EM='-6 1982 Sirrulation 37.2 54.6 1981 E OilJt14R EW-6 1982 Sirrulation 64 86.8 1a).2 127.3 1981 OILT-W< EW-6 1982 Sirrulation 63 92.1 129.7 152.3 1981 ECM) E}f'-6 1982 Sirrulation 46.1 1981 Opeconanics EW-6 1982 Sirrulation 39.7 41.5 1981 c \tOIL EM='~ 1982 Sirrulation 47.8 69.6 81.8 69.1 1981 r: Singerb Singer 1933 r-1ono opt 19.46 23.95 29.96 35.21 1981 lJ SI-CAC 9-I:A 1983 ~~ 23.79 24.51 29.26 33.93 45.59 58.35 1981 b Texacod Texaco l983 ~ 3.46 26.4 27.34 28.29 1981 Devarajan et al Devar. et al 1~ Sirruhtion lL93 1981 c llU '.a1 llU 1~ Sirrulation ]).09 l:i.l 44.12 51 1981 L E [ r [j [ [ Organization or r'odel ~ Hig, ~ LQ!i ().15 Mid 4 Soorce or r.txtel EIA EIA EIA [ RWIBSset et al RWIBS et al 51-f.AC 9£A [J ~td ()15 Lew (J (M) Mfd ~ ()15 Hig, [ ()15 LQ!i Draft c CM5 Mid Draft eMS Hig, Draft c ~ EIA EIA EIA · EIA EIA EIA 41 Table 3.2. (continued) Date of r-txte 1 Source T~ lS63 Sinulation lS63 Sinulation lS63 Sinulation 1~ sinulation ]g34 ~ ]g34 Sinulation 1$4. Sinulation 1.934 Sirrulation l934 Sinulation ]g34 Sinulation 1934 Sinulation 1~ SiTTUlation 1$5 32.(l; 19.8 Z3.57 17.12 Z3.79 ~.13 aJ.31 Z3.99 ZJ .62 2l.(l; ~.5 ai.33 ml 45.26 ai.4 34.88 19.56 25.25 21.94 26.39 33.17 41.3 21.94 26.33 35.11 aAs reported in Foster, Burton and Ha~ter ( 1933). ~ ~~c". ~" Year Constant 1995 am ana a::Ja:l Dollars 25.47 ~ ~.19 ~-~ n.~ 24.57 33.(l; 45.69 59.62 26.32 35.11 48.27 ~1 1$1 1~1 1$1 ~1 ~1 ~1 ~1 ~1 ~1 ~1 1981 e E bsinger jX)ints rut that these are oot oil price forecasts, M rather reflect tre prices that YGJld t:e charged !?( a profit ITBXimizing oil cartel. c r u 6 c C cAs reported in A 1 aska P<W!r Authority ( 1~) • drnterpreted fran a ~ical presentation. c. L [ [ .. Source 43 Table-3.3. (continued) Date of MJdel Year Constant Fiorganization U or r.tldel or MJdel Source T1E! 1985 ].9CJ) 1995 am ana aJ20 Do 11 ars PMITREb --'Fosterb [Sii'Ab MITRE Foster 9£.A 1$1 ~ 1$2 ~ 1$2 ~ 32.()) 35.83 :D.73 62.7 71.66 46.2 97.5 Sli'Ac 9£.A NO~ 23.79 24.51 29.26 33.93 45.59 58.35 Crexacoa BS!i'Ac ~lant IQ=' ['D1 . LJ Singerd [lA tlacro ['<~-Nehring Bdli Texaco 1$3 NA 9£.A 1$4 ~ E}tf'..(i 1$2 Mooo Opt tlarshalla 1$2 Moro Opt . Singer 1$3 rr1ono Opt a.f"..() 1$2 COfll Opt a.f"..() 1se2 Call> Opt Sohi 1332 Cootir~ation arnterpreted fran a graphical presentation. 25.46 23.79 29.73 19.46 39.41 0' e bAs reported in Foster, Burton and Harppeter ( 1983). cAs reported in Alaska Pa..er Authorit;y (1~). ai.4 'l/.34 28.29 25.25 29 36.19 54.29 72.39 55.5 71.3 ffi.8 1()).1 33.86 36.63 39.76 53.43 23.95 29.96 35.21 5),6 00.2 64.1 68.9 56.8 n.s 71.4 71.4 1331 1331 1331 1$1 ~1 ~1 1981 1981 1981 1981 ~1 ~1 h dsinger !))ints wt that ttlese are rot oil price forecasts, rut rather reflect the prices tilat VOJld re charged ~ a profit rraximizing oil cartel. c . F u 6 L [ [ [ .. 44 [ . Table 3.4. A SU11Tilry of oil price forecasts an-anged by t organization making forecast (In coostant 1~1 oollars) / Year [ Organization Soorce Date of r.'odel Constant or ~t>de 1 • or r-'ode 1 Soorce Type 1~ 19Cll 1.995 am ana ::.U20 Dollars [ B<tti Sohi 1932 Camination 39.41 1981 Devarajan et al Devar.et al ~ Sinulation ll.93 1981 [ aE;OPPAa OOEA:lPPA 1$1 ~ 43.37 51.86 69.77 1981 !:Rl 101 [Rl ~ Sinulation l).09 36.1 44.12 51 1$1. E [Rib [Rl 1934 Sinulation 24.13 21.94 24.57 1981 r: ETA ft4acro EJ.F-6 1932 C<Jtl) Opt Sl.6 00.2 64.1 68.9 1$1 (_; Fostsera Foster 1982· ~ 32.()) 35.83 46.2 1981 [ Gately a.f'-6 l982 Sinulation 52.9 71.7 71.3 93.8 1981 IPE EM='~ 1~2 Sinulation 37.2 54.6 1981 [ Kennectt-Nehri ng EM='~ 1~2 C<Jtl) Opt 56.8 77.5 71.4 71.4 1981 c MITRE a MITRE 1931 ~ 71.66 1981 OilJ."AA tw~ l982 Sirrulation 64 86.8 1ZD.2 127.3 1981 e OILT..aN< tw~ 1932 Sirrulation 63 92.1 129.7 152.3 1981 1981 (M) tw~ 1932 Sirrulation 46.1 t (M) Hi(jl EIA 1SW Sirrulation 32.()) 45.26 1981 (M) Hitjl EIA 1934 Sirrulation l/.62 41.3 59.62 1981 c CMS Hi(jl Draft EIA 1934 Sirrulation 26.33 35.11 48.27 1981 r~ CMS L~ . EIA 1.933 Sirrulation 19.8 a5.4 l981 u (M) L~ EIA 1934 Sirruhtion ::.U.31 a5.39 33.()5 1981 [ CMS L~ Draft EIA 1934 Sirrulation 21.()5 21.94 26.32 1981 CMS Mid EIA 1.933 SirTlllation 23.57 34.fe 1981 L CMS Mid EIA 1934 SirTlllation 23.99 33.17 45.69 1981 L CMS ~'i d Draft EIA 1934 Sirrulation 24.5 26.33 35.11 1981 [ [ L r:-.- 4 Soorce 45 Table 3.4. (contii1Jed) Date of t-b:iel Year Constant Organization cr~l LOPE OOE or r.tJdel Soorce T~ 1~ l.gg) 1995 am a:no a:J20 Do 11 ars [ OpecoriCJlrics Rrumsset et al [ Sa1ant ICF [l Satty & Ghol. L SH:Aa rs;[AC l.__; 51-[AC [ Singerd [T~r:P IDIL c}l)M CPE EJ.F-6 RCUTBS et a1 EJ.F-6 Satty & Ghol. 9-CA 9-CA 9-CA Singer Texaco EJ.F-6 tAarsha1la J3D NA 1~2 Sirru1ation 1~ Sirru1ation ~2 Mono Opt 1~1 Political ~2NA ~NA ~NA 1~ Mono Opt ~NA 1~2 Sirrulation 1~2 Mono Opt 47.02 39.7 17.12 19.56 41.5 25.47 55.5 71.3 88.8 1C6.1 4l.C6 44.86' ~.73 62.7 97.5 23.79 24.51 . 29.26 33.93 45.59 58.35 23.79 25.25 29 36.19 54.29 72.39 t9.46 23.95 29.96 35.21 25.46 26.4 ZJ.34 28.29 47.8 69.6 81.8 69.1 29.73 33.86 36.63 39.76 53.43 0 t aAs reported in Foster, 8urton and Hanpeter ( 1~). brnterpreted finln a graphical presentation. cAs reported in Alaska Pa...er Authorit;y (1~). 1~1 1981 1981 1981 1981 1981 1981 1981 1~1 1981 1981 1981 clsin~r ~:Qints rut that these are rot oil price forecasts, rut rather reflect the prices tilat v.oold oo ccha~ by a profit IT6Xinrizing oil cartel. F w b [ [ c [ 4 46 r Table 3.5: A Sllmllry of oil price forecasts amnged ~ LJ D projected l,gg) price level (In CDlStant 1~1 oollars) Year rOrgani zation Soorce Date of ~1 Constant or~l or r.txEl Srurce T:tpe ~ ].9g) 1995 a:m 2)10 2J20 Dollars cMITRE· MITRE ~1 NA 71.66 1$1 B<tti 1.Se2 Camination 39.41 1$1 -~B<tti [Devarajan et al Devar. et al 1983 Sirrulation 3).93 1$1 RQIIBSset et al RQIIBS et al 1983 Sirrulation 17.12 19.56 25.47 1$1 p~rb llU ~ Sirrulation 24.13 21.94 24.57 1$1 CMS L~ Draft EIA ~ Sirrulation 21.()) 21.94 26.32 1$1 [s;ng:rC Singer ~ MoroOpt 19.46 23.95 29.96 35.21 1$1 [9-f.Ad 9-rA 1983 ~ Z3.79 24.51 29.26 33.93 45.59 58.35 1$1 51-f.Ad 9-rA 1934 NA Z3.79 25.25 29 36.19 54.29 72.39 1981 [J.1S Mid Draft EIA 1934 Sirrulation 24.5 ~.33 35.11 1981 CMS L~ EIA 1934 Sirrulation 2J.31 ~.39 33.C6 1981 C~exacab Texaco 1~~ 25.46 ~.4 27.34 a3.29 1981 r:: EIA 1~ Sirrulation 19.8 ~.4 1981 EIA 1934 Sirrulation Z3.99 33.17 45.69 1981 [l)f M!rshalla 1~2 Mono Opt 29.73 33.86 36.63 39.76 53.43 1981 (}1S Mid EIA 1983 Sirrulation Z3.57 34.88 1~1 eMS Hiljl Draft EIA 1934 Sirrulation ~.33 35.11 48.27 1981 y5osterd Foster 1982 ~ 32.()) 35.83 46.2 1981 ~I 'Of au l983 Sirrul~tion 1).09 -36.1 44.12 51 1981 [iPE t}f"-6 1982 Sirrul1tion 37.2 54.6 1981 Opecooorri cs 8-F-6 1982 Sirru l-3tion ~.7 41.5 1981 L~ Hi~ EIA l£&l Sirru 1 at ion ll.62 41.3 59.62 1981 (atzy & Ghol. Satt;y & Gho 1. 1~1 Political 41.()5 44.86 1981 E [ ~ 47 [ Table 3.5. (cootirued) C Organization Year Source Date of r.tldel Constant or r.tldel or~l SaJrce Type 1S65 19.}) 1995 am ana aJaJ Dollars L (M) Hi!Jt EIA 1$3 Sinulation 32.(); 45.26 1981 [(M) EM="-6 1$2 Sinulation 46.1 1981 ' OPE rDE CPE ~NA 47.02 1981 [ WHL EM="-6 1~2 Sinulation 47.8 69.6 81.8 69.1 1331 ETA r-'acro EM="-6 1~2 Call) Opt 5).6 00.2 64.1 68.9 1981 c lll:AJPPAa rDEAJPPA 1$1 NA 43.37 51.86 69.n 1331 Gately EM="-6 1~2 Sinulation 52.9 71.7 71.3 93.8 1331 c Salant ICF t}f"-6 1~2 ~-1ono Opt 55.5 71.3 83.8 ll)).1 1331 . [ Kerlne<tf-Nehri ng fW-6 1~2 Call) Opt :x5.8 n.5 71.4 71.4 1981 Si-[Ad SI-CA 1~2 ~JA 5).73 62.7 ' 97.5 1981 [ OILT.AU< EM='-6 19:s2 Sinulation 63 92.1 129.7 152.3 1981 OilJAAA EM="-6 1982 Sinulation 64 ffi.8 laJ.2 127.3 1981 c ._ 0 aAs reported in Foster, 3urton and Ha~ter ( 1983}. brnteJ'l)reted fran a graphical presentation. r cs;nger !X)ints <lit that these ar not all price forecasts, rut rather reflect the prices that YOJld ~ C charged ~ a profit rraximizing oil cartel. c ~s reported in Alaska Power Auti1ority ( 1~}. f L [ L L 0 c [ [j c [ [ c r LJ [ [ c B b c F ld [j L [ 48 $33.06">, between $25.47 and $97.50 per barrel In 2000 (wIth a median of $62.10), between $45.59 and $129.70 per barrel In 2010 (with a median of $71.40), and between $53.43 and $152.30 per barrel In 2020 (with a medl an of $71.90). Caution must, however, be advIsed In I nterpretl ng these resuf ts. AI I forecasts do not give price projections for the same time per!ods. General fy, fewer projections are available for years further Into the future. Some general conclusions can, however, be drawn from the review of these forecasts. First, price forecasts do not seem to vary consistently according to model lng type. Projections vary significantly within the model lng types for which several forecasts are given. For example, projections for 1990 using simulation models range from a high of $64.00 per barrel for a 1982 forecast to a I ow of $19.56 per barrel for a 1 983 forecast. Further, forecasts published dur!l'"lg 1982 as a result of the Energy Modeling Forum's EMF-6 report en world oil prices--ir. which the s!mul~tlon approach was used--varied between $37.20 and $64.00 per barrel. It is interesting to note, however, that the forecasts based on monopol istlc optimization tend to be relatively lower--especlally in the rr:cre distant future--than forecasts usirg ccrn~etitive optimization and slmul~tlon models. The reader r.1c-y· recall fran the previous chapter tr.at economic theory suggests that although a monopoly w!f I set pr:ces higher than competition wculc dictate at the beg!nning of the production of the fixed resource, a monopoly or a cartel will eventLally set prrces lower thaf"l wculc exist if"l a competitive market as resources are depleted. n [ [ l C [ [ c r u [ [ c [ b [ F L [i 6 [ 49 4 Second, there does appear to be a slgnlffcant difference between forecasts publ fshed during the 1981-82 and 1983-84 time periods. A rev_few of Tables 3.2 and 3.5 shows that forecasts have In general been revised downward In the 1983-84 tfme period. Table 3.4 shows how forecasts from specfffc models have changed over tfme. For example, base projections publ !shed by Data Resources, Inc. In 1984 were much lower than those publ !shed In 1983 --from $44.12 down to $24.57 dol Iars per barrel for 1995. Forecasts publIshed by the Energy Information Administration (within the U.S. Department of Energy> using the 0!1 Market Simulation <OMS) Model have been of 1 983. $34.88 per revised downward since the official. OPEC price reduction EIA's medium case projections for 1990 have changed from 7 barrel In 1983 to $26.33 per barrel In 1984. The high and low EIA forecasts have also-been revised downward from $45.26 to $35.11 per barrel and from $26.40 to $21.94 per barrel, respectIvely. As can be seen from Table 3.4, projections by Sherman H. Clark Associates (SHCA) have been lowered drastlcal ly since 1982. For example, SHCA projected In 1982 that ol I prrces would be $97.50 per barrel In 2000. In 1984 that projection has been revised downward to $36.19 per barrel in constant 1981 dollars. ~ote that the ~ost recent EIA forecast given In the tables Is a draft and thus cannot be consrcered an offlc!al EIA forecast. [ [ [ [ [ [' [ c r ~ [ [ c [ b c f LJ 5 L L 50 s. Sensitivities of Forecasts to Changes In Key Parameters There Is not sufficient detail given In the above d!scussed forecasts to c~plete a ·formal sensitivity study of the results. Ideally, one would like to test the sensitivities of different model lng types to changes In the key parameters. However, In most cases the publ !shed forecasts give only minimal Information about the key parameter values. Many parameter values are not given. Further, In many publ !cations there are several scenarios presented; however, the different scenarios represent changes In several parameters, not just one. It Is therefore not possible to determine how the forecasts vary because of a change In a part I cuI ar parameter. As a result, this subsection wll I concentrate only on the publ !cations that have reported how forecasts from a particular model or a group of models vary due to changes In the major parameters. Two publications are the focus cf this subsection the Energy Model lng Forum's EM="-6 report and a book edited by Griffin and Teece (1982). The discussion Is further limited to the majcr parameters that Influence the forecasts of optimization and simulation models d!scussed In Chapter 2 --i.e., ol I cemand price elasticity, GNP growth, o!l cemand income elasticity, Interest rates, the backstop price, and the level of CPEC production capacity. Table 3.6 gives the different scenar!cs used in the E~1F- 6 st~.;dy. Figure 3.1 sl.ll1mar~zes how the ell price forecasts for 2000 from each of the ten simulation and optlmizatiqn models used In that study var:ed, frven the ma~cr parameter changes as repres~nted by the c: tferent scenarrcs. c 51 c 4 Table 3.6 Scenario ~ons used in the EW-6 s111ctf ....; [ Dal8ld L~run Eccnmfc CFEC Scenario RecU:tf~ Prillllry Ener'qf Grotth Proctlction Noncalvet'ltiona 1 Oerand Elast1c1zy Rate capac;zy Energy $t.q) ly c 1. Referen:e None 011: -0.6, See 34 r+1ll $60/bb 1 : 1 i mf ted erergy: -0.4 Table A-3 CJ,Jantities [ {Appendix B) 2. Oil Deland 2) f+8) [ ReclJctlon by:ml 3. LQII De1and ~.375, [ Elast1c1zy ~.25 4. Oil Oel1llnd ReclJctlm-c LQII Oe1a1d 2) f+8) -0.375, Elast1c1zy by 2)3) -0.25 r 5. LQII Ecancmfc 2/3of Gmrrth referera L rates [ 6. Restricted m of refenn::e Backstcp ' limits [ 7. Disrwtion ~ r+1ll fran 1~ on c 8. Technological $40/bb 1 ; i ocreased Breakthf'Cll41 limits l 9. Disruption-~ t+Bl LQII Dellll1d ~.375, fran Elasticizy ~.25 1985 on c 10. Optinrist1c aJ r+Bl ~<ilal $40/bb 1 ; increased by2l3l ircrease limits c 40 t+Bl in 1987 11. Disrwtion-aJ ftMJ) 24 t+BJ r 0 i1 DE!IIV1d by ~aJ fran u ReclJctlon 1$5 on t 12. Hi~ Oil Price c SOJrce: EJof-6 (1982, page 102) [ [ [ r L c [ [ [ C r L [ [ c E c c f u t ~ [ ,. sceaar1o 1. Reference 2. 011 DU&n4 Reductian 3. Lav Deulld Ehat1c1ty 4. 011 Deulld Recluotian- Low D-£laat1c1tY 5. Low Econo&I.C GI'Oiftb 6. _Rest~1cted Baclcatop 7. D1a~upt1on 8. TeohiiOlOClcal B~eal<tMoucb 9, D1a~upt1oa-Lov Deund Elaat1o1ty 10. Opt1&1at1c 11. D1arupt1on-01l • Deund Reduction 0 20 c B 52 wr,~~~o~~lr~;·~~ ~.f~OO -0 B 8 60 80 CGI\ISl A I \1 A 0 & G S IC W I G \1 I C A S G 0 ! 8 I C 0 AI G W S IC B G c B s111 a W S A I l C W I S l, A S I A w l 0 W G I& ! 0 B G\1 lKA Of! c s 100 s ' 0 120 £ 0 1C s 11oaela: G : Gately, I • lEES-OMS, C • IPE, A • ETA-MACRO, IC • Konnedy-Neftr1nc, 0 • OILMAR, E: OILTAIIIC, W • \IOIL, S • Sahnt-ICF, 8 • Opecono.ic.s 1-0 160 c £ & 0 A Note: For all IIIQC1ela otner tnan tEES-OMS and IPE, the ave~ap or prtcea oetw"n 1995 and 2005 1s given. For IEES-OHS, the 1995 prt ce u presented; ror IPE, averages between 1995 and 2000 are presented. Several projections are nigher t.nan 1160/bbl anel thu.3 do not appear above. These include: ror the low deund elut1c1ty scenario, Kennody-Heftrtnc ($175) and OILM.U ($177); ror the aurupUoa-low <1-.!d elut1c1ty acenar1o, OILTAJIIC ($184), IPI ($198), hnnody-Neftr1nc ($217), &na OIUIAR ($-17). Fig. 3.1. Sensitivities of EMF-6 price forecasts. Source: EMF-6 (1982, page 49) c [ [ c [ [ [ c r L [ [ c c [ c r u [ [ [ 53 1. Long-Run 011 Prtce Elasticity of Demand There has long been a controversy about the appropriate long- run oil price ~lastlclty· to use In oil price forecasts. In the EMF-6 study a base elasticity of -0.6 was used. In other scenarios the price elasticity was reduced to -0.375. As can be seen from Figure 3.1. this change resulted, as we might expect, In large Increases In the forecasts for the year 2000. In the base case. prices range from about $40 to $95 per barrel In constant 1981 dol Iars. The price elasticity reduction Increases that range to about $45 to $145 per barrel. The base price elasticity used In most papers In Griffin and Teece <1982) Is significantly higher at about -0.75. Daly, Griffin, and Steele <1982) study the lmpl fcatlons of a -0.365 demand elasticity on OPEC's stabll lty and conclude that under the lower elasticity --and given a $32 real price path --OPEC's production Increases to 38.3 mil I Jon barrels per day In 2000 as compared to 22.2 at the higher elasticity. OPEC's stab! I lty Is thus more stable at the lower elasticity and therefore higher oil price paths are more I lkely. In a sensitivity study by Salant (1982) the pr!ce elasticity of demand was reduced from -0.5 to -0.4 with the result that the optimal price for a profit maximizing cartel Increases by 7.6% In 1980. The price difference between the two paths Increases gradually over the entire forecasted period --1980 to 2050 -- given the elasticity reduction. c [ [ D C [ [ [ r L [ L c c ( c r u B L [ 54 2 ~ G~P Growth There appears to be less dtsagreement about the ass~.med rate of GNP growth. T~e EMF-6 study assumed annual GNP growth rates of approximately 3J In the OECO countrres and 5J In the oll-tmportlng developing countries through the year 2000. Other forecasts used approximately the same rates. For example, Data Resources, Inc. used a GNP growth rate of about 3% for seven of the large OECO countries tn Its Autumn 1984 publ tcatlon. The Energy Information Admtnlstratron also used a 3% growth ftgure rn Its May 1984 pub I tcatl on. The papers In Grlfftn and Teece (1982) assume OECO economic actlvtty wll I grow at between 3J and 3.5% between now and the year 2000. During recent decades real GNP growth has been about 4%, with stgnlflcantly faster growth tn the 1960s as compared to the 1970s. There may be more divergence on how GNP growth wll I translate Into tncreased o!l consumption. In most of the models used In the EMF-6 study, It was assumed that growth In oil demand would be roughly proportional to the growth In GNP. In the papers In Griffin and Teece !twas assumed that roughly a 1% Increase In 8 GNP wll I result In a 0.75% Increase In oil demand. In the EMF-6 study .the base case economic growth was reduced by 33% In one scenario. Flgure 3.1 Indicates that most of the 8 The lmpl !cations of economic growth wll I differ depending on the type of country In which the growth occurs. Most forecast that growth In less developed countries wll I result In larger Increases in oil demand than In developed countries. For more on this subject, see the EMF-6 report or Griffin and Teece (1982>. fj [ [ D [ [ [ c r L [ [ c c E c F: lJ Q c [ 55 forecasts were moderately sensitive to the lower growth 9 ass1.111ptl on. 3. 011 Supply Prtce El~sttclty Balder (1961) cont~tns ~ summ~ry of the OECD otl supply price elasttctttes contained In the models that were a part of the EMF-6 study. That summary Is reproduced tn Table 3.7. Elasticities ranged from -0.241 to 1 .162. Balder concludes that the differences In the ass1.111ed supply price elasticities had a significant Impact on the EMF-6 forecasts. In Grtffln and Teece (1962>, which generally contains lower prtce projecttons than the EMF-6 study, the authors conclude that "<t>he prtnctpal differences between the vtews presented here and the Stanford's Energy Model tng Forum appear to stem from assumptlons about OPEC's behavtor and the non-OPEC supply response. · ••• The dffferences are not principally due to the prtce elasttclty of demand assunptlons ••• " <page 214). In addition to bel !ev!ng that the supply poss!bl I !ties In the non-OPEC countries are greater than the EMF-6 models assume, the papers In Griffin and Teece argue that the assumed OPEC production capacity of 34 m!l I ton barrels per day Is "exces!vely conservative." "(l)f OPEC production reached 38.3 (million barrels per day) In the year 2000. reserves to 9 Balder (1981) contains an analysts of why the different models In the EMF-6 project produced different price forecasts given the same Input assumptions. Seider concludes that alternative assunptlons about non-OPEC supply price responsiveness greatly Influence the divergence In price forecasts. Balder also concludes that "(t)he Inclusion of a feedback effect, In which higher of I prices reduce the economic growth of the of I Importing nations. Is significant In moderating the magnitude of price changes" <page 3). [ [ 4 56 [~ Table 3._7. \£CO oil ~ly elasticities: DF-6 roodels Q roitls llB; ].gg) 1995 ~ ani 2)10 1!)15 1!)3) Average [ IEES,.a.tS (1) 0.026 o.3n 0.455 0.283 (2) 0.033 0.319 0.197 0.183 (3) 0.(!;6 0.266 0.425 0.249 [ IPE (1) 0.0 0.0 0.157 O.CBJ 0.060 (2) O.OZ3 0.046 0.119 o.~ 0.()32 (3) 0.017 0.1()) 0.042 0.1ai 0.068 [ Salant/ICF (1) -3.591 ~.~ 0.115 0.509 -2.814 4.7Z3 2.049 2.225 0.3>2 (2) -1.540 -!.all ~.621 ~.863 0.284 0.718 0.552 0.742 ~.241 0 ETA~ (1) 0.713 0.439 0.271 0.569 1.911 0.954 2.628 1.745 1.154 (2) 0.0 1.114 1.521 0.317 co s.n 1.129 -1.251 1.162 (3) 0.0 0.232 0.561 0.876 0.746 0.001 o.n9 0.788 0.5~ r WJIL* {l) 0.007 0.035 0.157 0.318 0.529 1.107 4.402 -2.~ 0.457 u (2) 0.009 0.027 0.200 0.166 0.392 0.468 0.604 0.371 o.m [' Kerlle<:tf I (1) 0.009 1.n 0.542 0.841 -1.053 co co co O.S<li Nehring (2) ~.191 ~.194 ~.193 ~.099 o.cm 0.318 0.~ 0.245 0.029 [ OILT.e.N< (1) O.<li2 1.568 0.858 0.358 ~.005 ~.270 ~.484 ~.4(!; 0.199 (2) 0.037 o.m O.SfB 0.103 ~.101 ~.253 ~.257 ~.175 0.042 (3) 0.114 1.101 0.599 0.047 ~.:m ~.483 ~.491 ~.JJ7 0.035 c Opeconanics (1) 0.0 0.0 0.0 0.0 0.0 (2) 0.0 0.0 0.0 0.0 0.0 U ortJ.M* (1) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 (2) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 c Notes: ( 1) = Reference Ri Oil DSII!M Red.lction s:enari OS (2) = Reference R1 Lew Elastici~ scenarios c (3) =Reference Ri Higt Price scenarios * =u.s. elasticities co = coostant prices ll.rt: a rmzero SLWlY dffferen:e between scenarios; f these cases are rot i rw: 1 u<Ed in tile averages u SaJrce: Bei<Er (1~1, page 28) 0 c . l c [ [ c C [ 1 -~ -J [ [ L [ D c [ [ c r u ~ c [ 57 produ~tlon ratios In the Cartel Core would not be appreciably different than In 1980" (page 215>. In the EMF-6 study It was shown that the stmulatlon.models structured around an OPEC capacity uti I lzatlon rule gave significantly lower price forecasts when the capac! ty I eve I was I ncr eased from 3 4 to 40 m I I I I on barrels per day. See Fl gure 3 .1. 4. Interest Rates The assumed Interest rate or discount rate can--as discussed In Chapter 2 --have a significant Impact on the optimal prtce level set by a proftt maxlmtzfng producer of a'depletable resource. In a paper by Marshal Ia and Nesbttt <1984> the sensitivity of changing the discount rate on tne opttmal price path of a profit maxtmlztng o!l _cartel Is assessed. In the case of this parttcular study, a change tn the assumed real discount rate altered the optimal price path less In more distant time periods than fn the ; near term. Under the assumption that the cartel ts composed of alI OPEC countries, a decrease In the discount rate from 6% to 2% Increases the optimal cartel price from $26.33 to $36.25 per barrel In 1987, from $32.42 to $40.37 per barrel In 1997, and from $49.36 to $50.63 per barrel In 2022. 5. The Backstop Price The EMF-6 study contained one scenario In which a perfect substitute for crude oil becomes available In the year 1996 at a cost of $48 per barrel In 1981 dol Iars --see scenario 8 In Figure 3.1. As can be seen tram F!gure 3.1, this particular technological [ [ r r: [ r [ c r L [ [ c c c G r b [ c L 58 4 breakthrough dfd not result In sharply different price reductions from either the stmulatton or optlmlzatlon models. IV. Surveys of Ofl Prtce Forecasts Recently an Workshop ClEW> date, long-term Informal group known as the lnternattonal Energy was formed to collect and compare the most up-to- forecasts of world oll market trends. The group sends out perlodlc surveys to numerous contributors who provide stattstlcs on crude-or I prices, GNP growth, prfmary energy consumption and production, and electrlcfty generation. The respondents do not provide Information about their model lng approach, the basic underlying assumpttons of their forecasts, and so forth. The surveys do, however, obtain responses from all sectors that forecast market condttlons --e.g., government agencies, private firms, academia, Individual energy consultants, and world organizations. Numerous countries are represented. Typically, the contributors provide only a base case scenario; however, In some cases contributors wll I provide, for example, high, medium, and low cases. Projections are given for the years 1990. 2000. and 2010. In this section the results of two recent lEW surveys are reviewed. The results of the 1983 survey are given In a paper by Manne and Schrattenholzer (1984). The results of a more recent lEW pol I completed In July 1984 were given In a presentation by Alan Manne at the November 1984. meetings of the International Association of Energy Economists. Table 3.8 contains summaries of those surveys. [ c [ l c [ ,, L. 0 r l.__; [ c c r L G c F u r~ c c L 59 In the lEW surveys alI price proJections are given In terms of Index nunbers where 1980 equals to 100. Note that In Table 3.8 the lEW results have been converted to constant 1980 dol Iars per barrel for Saudi Arabian I lght crude. The CIA (1984> reports that the average prfce for Saudi Arabian I lght crude In 1980 was $28.67 per barrel. As can be seen from the table, the median and mean forecasts have been revised downward from 1983 to 1984 by about $3 to $4 per barrel for alI forecast years. According to the latest JEW survey, the median forecasts for 1990, 2000, and 2010 are $27.81, $36.73, and $46.66. respectively, fn real 1980-dol Iars. The range of forecasts continues to be high, with-some forecasters predicting drastic real price reductions and some predicting drastic price rises. c. [ [ l c [ [ u r L [ [ c c c c C r: c k [ 60 Table 3.8. Results of the 1983 and 1984 IEW surveys NUf.eER MEDIAN ME~ STANDARD RANGE OF RESPONSE RESPONSE DEVIATION RESPONSES FORECAST YEAR 1983 SURVEY 1990 68 $31.54 $31.90 $7.94 $63.93-$20.34 2000 61 $42.43 $41.28 $10.55 $68.80-$17.77 2010 24 $50.17 $49.63 $14.74 $76.26-$14.34 FORECAST YEAR 1984 SURVEY t 990 57 $27.81 $29.11 $7.69 $63.93-$20.07 2000 54 $36.7 3 $37.69 $9.92 $62.50-St 8.64 2010 16 $46.66 $47.54 $15.74 $76.26-$15.77 [ [ [ l c [ [ D r L-> [ c c c [ c f) li [~· -. c L I • 6 t 4 4. CONQUS IONS "Energy forecasting Is a hazardous occupatfon. projection Js doomed to be Incorrect" Schrattenhofzer, 1984, page 48). Introduction V f rtua I ly arry (Manne and The purposes of this paper have thus far been to (1) review the different model lng approaches that have been developed to explain historical oJI price changes and forecast future price trends, and <2> review, Interpret, and criticize price forecasts that have resulted from these methodologies. In this concluding chapter we review the conclusions that can be drawn from Chapters 1, 2, and 3 In regard to these general purposes and address the degree to which there exists a consensus on model lng approaches and price forecasts. In addition, this chapter presents a subjective assessment of future all market trends based on the conclusions drawn from previous chapters. Of particular Interest fs· how market structural changes --resulting from both economic and pol !tical pressures--mav Impact future all prices. Finally--and again drawing on conclusions from previous chapters --this chapter contains thfs author's subjective judgment of all price trends for the 1987 to 2022 tfme period. In addition to providing high, medfum~ and low price trajectories for "business-as-usual" market conditions--conditions In which the basic structure of the world all market remains relatively unchanged--the lmpl !cations of a severe oil supply disruption and a ruther breakdown of OPEC's cohesion on oil prices are also considered. [ [ [ r C [ r c r L.J [ [ c 0 [ c F u 6 L [ 62 I 1. A Summary of Conclusions from Previous Chapters A revtew of the conclusions from the previous chapters Is In some WtlfS dtfftcult., because one of the main thrusts of those chapters Is that after more than a decade of Intensive study of the world oil market there Is I lttle consensus about the way prices are formed--which Is., of course, In Itself a major conclusion. However., In a more positive vein there are several general conclusions that can be drawn from a review of the hlstory of the oil market., the models to represent that market., and the forecasts from those models. The overrldtng concluston Is that ln alI three are~s --I.e • ., the market., the models, and the forecasts --there has been slgnlflcant evolutlon within the past decade. A review of the structural changes In the oil market slnce the late 1960s shows drastlc and contlnual changes tn the relevant players that have slgnlflcant control over the pricing mechanism, the relevant objectives of those players, and the abl I !ties of those players to manipulate the market to real lze those objectives. During the 1970s the major producing countries gained significant control over production decisions within their own countries and thus gained sIgnIfIcant control over oil prIces. In the ear I y_ 1 980s that control began to decl lne as non-OPEC producers Increased production and maJor consuming countries reduced consumption In response to the drastic price escalations of the 1970s. In addition, the objectives of the players--particularly, but not exclusively, the major producers --rr.ety have changed over time as pol !tical objectives outwel~hea wealth maxlmlz!ng goals and as demands for 0 c [ t [ [ [ 0 r: L..; [ [ 0 u G c F bJ l c L 63 oil revenues became tncreas!ngly Important to ambitious development progrsns. Pol Jtlcal decisions within some consuming countries also led to structu~al changes--for example, the removal of oil price controls In the United States. There has been an equally Impressive evolution In the ways the market has been modeled. During the early 1970s models largely followed dlsclpl lnary lines, with variations within each dlsclpl tne to account for different perceptions of the structure of the oil market. To a great extent these models were built by economists who typically used very rigid orthoqox theoretical approaches to represent a market that was, In fact, undergoing great structural change. Very elaborate mathematical models arose to represent how a monopol Istre producer would produce or set prices to maximize discounted profits. However, the sophistication was based largely on tne theoretical and empirical refinement of a rather slmpl tstlc underlying objective function--wealth maxlmlzatton. Economists In large part Ignored the fact that major oll producing countries are not subject to the same kind of forces that a competitive or even monopol lstlc producer faces wlthln a capital lstlc market environment--I.e., the threat of takeover when profits are not maximized. From this real lzatlon arose what may be termed "sufficing" models that adopted alternative objectives, such as meeting a minimal revenue requirement. The decision to employ producer objectives other than wealth maximization resulted In models that more closely resembled the real-world ell market; however, It also opened a whole new set of Issues. If profit maximization Is not necessarily an objective u [ r l_) u c [ [ c r L~ [ [ 0 0 [ c f u lj L L 64 4 that the major producers wll I follow, or wJI I eventually be forced to follow because of market pressure, what objective should be used? It Is ot this point that many modelers real lzed that pol ltlcal objectives plav a major role In the production and prJcJng decisions of many producers and should be expl lc!tly or Implicitly Inc I uded In any model of the world oil market. The wavs pol !tical considerations were Included differed. In sane models pol !tical constraints were Imposed on an otherwise economic wealth maximizing approach. In other models the pol ltlcal process was tantamount to or exceeded alI economic market forces. Some modelers opted for Informal models that Jmpl lcltly contained economic and pol ltlcal considerations. Therefore the major conclusions to be drawn from more than a decade. of model Jng the world oJI market Is that the oJI market Js continually evolving and world oJI models must continually evolve to more closely represent that market. One can also conclude that the general thrust of future models wll I be to employ an lnterdlsclpl Jnary model lng approach to address the complex economic-pol !tical questions or to use a slmpl Istlc rule of thumb In forecasting prices, such as setting prices as a function of an exogenously specified OPEC capacity utJI lzation level. In addition, there Is an Increasing real lzatlon that any forecast of world ell prices Is highly uncertain. The general trend Is to deemphasize specific point projections and concentrate more on the probable range of future prices. Although recent forecasts of oil prices for the next 30 years are generally down by· $3 to $5, reflecting the recent official OPEC price reduction, the r: L;. [ [ t [ [ [ c r LJ [ [ 0 6 c c r; L [i C L 65 vartatton tn the forecasts rematns high. The medtan response of recent forecasts Is that Jn real terms oil prices wll I remain flat until about 1990. Be~een 1990 and 2010 the median forecast Is that real prices wll I Increase by between 2 and 3 percent yearly. J I I I. A Subjective Assessment of Future Market Trends Given the various caveats discussed In the above section, this sectton dtscusses the current oil market environment and presents thts author's own subjective assessment of future market cond I tl ons. At the time of this wrlttng the oil market continues to be "soft" and many Industry offtctals predtct further price cuts. In an effort to support Its $29.00 base price, OPEC recently agreed to cut Its production ceiling from 17.5 to 16 mtl I ton barrels per day, with the major producing countries accepting the majority of the cuts. However, otl compantes continue to put pressure on OPEC and non-OPEC producers to make prtce concessions as of.ftctal prices continue to exceed prices on the spot market. For example, pressures on producers of North Sea oil --!.e., Norway and the United Kingdom--place Indirect pressure on Nigeria which produces crude of a stmllar type. Nigeria's wei I publ lcfzed revenue needs make price concessions more probable In the event of a North Sea prtce reduction and thus continues to threaten the prtctng structure of OPEC. However, one must be careful .to distinguish between short-term market signals and probable long-term market posslbll ftles. In the long term three widely divergent price paths are possible. First, the all producl ng countr f es may "weather the current storm" and maintain their pricing cohesion, It not their current $29.00 base [ [ [ u [ [ [ c r Li [ [ c c [ c f: u b L [ 66 4 price. The strength of OPEC Ts largely dependent on the abl llty of Saudi Arabia to absorb the necessary production cuts to maintain their desired price. Second, demands for revenues In the major producing countries mav force production levels that are Incompatible with current prices, or the major producers may be forced to "pol Ice" members of OPEC and non-OPEC producers by allowing prices to fal I, at least temporarily. A significant price reduction of this type could alter the world price trajectory for a period of decades. Third, pol ltcal turmoil In the Persian Gulf area that results Tn the long-term disruption of crude from Saudi Arabia and other large Persian Gulf producers could cause drastic escalations In price, such as those observed In the 1970s. Most oil analysts agree that such a disruption, even during soft-market conditions, would result In severe price Increases. The most probable of these three broad scenarios Is the first I.e., the present basic structure of the market wll I remain Intact. It Is generally felt t'hat the benefits to be received by oil producers from maintaining the current structure are large enough to prevent OPEC and non-OPEC producers from entering a "price war" that could significantly reverse the price escalations of the 1970s. Over the longer term, It Is general Jy agreed that OPEC wll I Increase Its share of the ol I market as non-OPEC reserves dwindle and world economic growth--especially In the developing countries causes significant Increases In world ol I consumption. Financial reserves within the major producing countries, such as Saudi Arabia, should be sufficient to allow production reductions that will prevent an "all-out pr!ce war." n [ [ r: [ [ [ c r: L [ [ c c L c F ld ~ b L 67 ~Is does not mean, however, that the probabll ltles of the other two extreme scenarios are Infinitesimally smal I. In another paper by this author <Curlee, 1984), It Is argued that the current pressures being exerted on the structure of the world oil market Increase the probabilities of both further, and possibly severe, oil price reductions and severe price Increases, as compared to the more stable market conditions of recent years. On the one hand, revenue pressures In the more ·populous producing countries may lead to pol ltlcal unrest that may result In the current governments of those countries being replaced. An example of these pol ltlcal pressures Is the recent coup In Nigeria. In order to Increase revenues, some countries within OPEC may elect to Increase production with or without the approval of OPEC. There Is evidence that this Is happening currently. If this occurs, more pressure wll I be placed on the major producing countries to reduce their levels of production. There are, however, I lmlts below which even the major producing countries cannot be expected to reduce production. Chase Manhattan Bank (1984) projects that OPEC's current-account deficit will Increase to $16 billion In 1985 fran a level of $15 bl II ton In 1984. Although the major core members of OPEC--I.e., Saudi Arabia, Kuwait, United Arab Emirates, and Qatar --have about $300 bl II ton In foreign assets, It Is clear that they could not be drawn down significantly without some degree of Internal turmoil. In the event that the core OPEC members no longer pol Ice their production I lmlts, severe price reductions are possible. These dow~ward pressures on price would be further Increased If Iran and Iraq end their war and resume production at E [ [ c [ [ [ c r LJ [ [ 0 t E c r u [ c [ 68 pre-war rates. OPEC as a whole has the ab!l Jty to Increase production by about 8 mrl I Jon barrels per day over their June 1984 10 levels, given their current available production capacity. On the other hand, the Internal pol !tical pressures resulting from unmet revenue needs could result In violent confl let within· and among the producing countries. Any v!olent confrontation Involving the maJor producers of the 'Persian Gulf area would probably result In severe price Increases. In the event of the loss of all Persian Gulf product!on capacity, total OPEC capacity would be reduced from 34.840 to 11.725 mil lion barrels per day. Production losses would be about 11.630 mil lion barrels, based on June 1984 production levels. Given that with the loss of the Persian Gulf producers, OPEC excess-production capacity Is only about 3.220 m!l I Jon barrels per day and I Jttle excess capacity exists In non-OPEC countries with the possible exceptions of Mexico and Canada--significant panic and price Increases could be expected In the world oil market. The exact price Increases that would result are virtually Impossible to predict. ThIs type of disturbance represents a worst case scenario. Of course, con~luding that the most I lkely scenario .Is one In which there Is no drasttc change In market structure only marginally reduces the problem of forecasting prices. Within this broad scenario there are numerous unknowns that may greatly Impact oll prices. However, drawing upon the opinions of numerous Industry experts, sane general cone I us! ons can be drawn. Fl rst, 10 The CIA (1984) reports that as of September 1984 total available OPEC produc-tion capacity was 26.175 million barrels per day. Production In June 1984 was 18.215 million barrels per day. c [ [ t [ [ [ c r L [ [ c 0 c c r~ 6 [J b c 69 there·ls significant evidence that neither non-OPEC suppl lers nor major consuming countries have fully adjusted to the drastic price Increases of the 1970s. · This continued adjustment wll I tend to dampen the demand for OPEC oil, at least throughout the remainder of the 1980s. Second, estimates of economic growth are relatively low which suggests that major off demand Increases wfl I not result from Increased economic activity. However, even If growth exceeds the current projections. recent work by Daly. Griffin, and Steele (1982> suggests a long-run ofl demand elasticity of only about 0.75 with respect to economic actfvtty. A key unknown, however, Is how econom rc growth In devel opt ng countrt"es w II I trans I ate Into Increases In oil consumption. Third, the level of OPEC production capacity does not appear to be a binding constraint In either the mid or long term. Huge levels of excess capacity currently exist and there appear to be few reasons that OPEC could not, or would not, Increase production capacity should the need arise In the long term. Fourth, various experts suggest that It Is not In the Interest of OPEC to desire large price Increase because of the I lkely responses of non-OPEC suppl lers and oil consumers to those Increases. Because of these reasons It Is most I lkely that real oil prices wll I remain constant throughout the remainder of the current decade. However, due to resource depletion In non-OPEC countries and In some OPEC countries, real price Increases are Inevitable eventually. Further, once the world or I market becomes "tighter" In the late 1980s and early 1990s, there Is an Increasing probabl I tty of short-term price spikes resulting from temporary c [ I L F b c [ [ c r L [ [ c l [ c f u [ b [ 70 suppl4y dt sruptt ons. There rs a growtng consensus. however. that prtce sptkes resulttng from temporary supply dl sruptfons wl II not result In drast!cally different long-run price trends because those hfgher prtces are not In the Interest of the major producers. Taktng the opinions of the various experts tnto account. real prtce Increases of about 2 percent per year appear to be most probable. IV. Future 011 Prtces Figure 4.1 gtves oJI prtce forecasts for the t[me pertod 1987 to 2022 In ftve ye~r tntervals under the assumption that there are no stgnlftcant market disturbances. The base case forecast t s computed under the assumptton that real otl prtces rematn constant tn 1984 dol Jars throughout the rematnder of the decade. Prtces for the remaInder of the · forecast pert od are assumed to t ncr ease by 2 percent per year tn real terms. The htgh and low projections are. admtttedly. reached tn an ad hoc way. If selecting a most I tkely base case from the numerous forecasts dtscussed tn Chapter 3 ts rtsky. selecttng a probable range around that base ts dangerous. In order to share that danger --or at mlntmum have some stated reason for selecttng a range--the standard devlattons of the forecasts from the most recent lnternattonal Energy Workshop were used. The htgh projections for the years 1987 and 1992 were obtained by addtng one standard deviation from the 1990 lEW forecasts to the base case results. The low projecttons were obtatned by subtracting one standard deviation from the base case. The high and low projections for the years 1997 and 2002 were obtained by adding and subtracting one standard deviation from the [J 71 [ ~ [ c y Table 4.1. Oil price projections 1987 to 2022 (In constant 1984 dollars per barrel) c Year Base Low High [ 1987 $29.00 $21.00 $37.00 [ 1992 $30.00 $22.00 $38.00 1997 $33.00 $23.00 $43.00 c 2002 $37.00 $27.00 $47.00 2007 $41.00 $25.00 $56.00 r~ L 2012 $45.00 $29.00 $61.00 [ 2017 $49.00 $34.00 $65.00 2022 $55.00 $39.00 $70.00 [ Note: Rounded to the nearest dollar c E [ c r u [ L - L c [ [ 6 [ [ [ c r l__; [ [ c C L c r u [ c [ 72 forecasts for the year 2000. AI I remaining high and low projections were obtained using the standard deviation from the lEW projections for_the year ·2010. It Is all but Impossible to project how prices might change under the two extreme scenarios discussed Jn the above section. In the case of a "price war" Jn which OPEC countries Increase production sharply In an attempt to Increase revenues, world prices could drop sharply--Into the $15 to $20 prfce range. Possibly the more Important question for consumers and producers wfthfn the Unfted States fs, however, the degree to whlc~ the U.S. government would allow prfces to drop. It fs most probable that a tarfff or quota would be set so that U.S. Investments In off production and use technologies that depend on relatively high off prices would not become Jnsfantly outdated. The tarfff or quota would be done In the name of --and could be argued for on the basts of reducing future U.S. vulnerablltly to events In the world off market. Real domestic prices below the $21 per barrel level might face significant opposition. In 1984 dol Iars the world price of oil was about $21.75 before the price escalation following the Iranian Revolution. Likewise Jt Is very difficult to predict how the market would react to a large and prolonged disruption of Persian Gulf ofl. In the EMF-6 report the models used In that study were exercised under the assumption that a permanent but unanticipated reduction In OPEC capacity of 10 million barrels per day would be Initiated In 1985 while the world ot I market Is projected to be "soft." Figure 4.1 shows how the varrous models responded to the scenario. For each c . ' [ [ c [ [ [ E ..,o; r u: -"' a:~ ~- L .. ...JQ. -"' 0-"' w= [ 0 0 ::>"0 a:-u~ [ c E [ c f u B G [ .. 73 ···••·•••·••••••••·•••· ETA·MACRO (\ ---Gately --------IEES·OMS f. \ · · · · · · · · · · · · Kennedv·Nehring .• \ ----OILMAR : •·.••• , OILTANK •. •· •••••. ,... ..... ·-·-·-~ -·-·-·-·-·• WOIL 7····.. '~-.... / . ··.. ' ..... /· ! ······... "·......... / I ··... ,~·, . ··.. . .... ;(. I .. • . ··... . .... . '• I I ·····... ..._/ ...... } : ···i; ·· .. .A.. ..... / ,' ; " -.. / : ,' i I "",, / .:' . .......... / :,' r ~~--...-·~ -----j· I i ,;' -............ ./.:.-'· ................. . 1/ ................. ... ,: ;:: / ...... ·: ,"" , -........ _ 120 80 ,. / 40 ---,.,,., ............... __ .,..,.,. 1982 1984 1986 19!!8 !990 1992 Fig. 4.1. EMF-6 price projections given a permanent 10 million barrel per day capacity reduction. Source: EMF-6 (1982, page 58) c [ [ ~ [ [ [ c r L [ [ c c E c F lJ [ ~ L 74 .. one million barrel per day reduction in capacity, the models -- with the exception of one--projected that prices would increase by between SS and $12 dollars per barrel. As can be seen from the figure, the impacts of the disruption significantly from model to model. scenario represents a worse case with occurrence. V. Final Conclusions over the 1 eng term vary Once again, this extreme a small probability of The dominant conclusions from this paper concern not what we know about the functioning of the world oil market, but rather what we do not know and how what we do not know can influence our perceptions of future market trends. In some senses, we as students of oil markets have asked as many questions as we have answered about how oil prices are formed. Further, there is an increasing recognition that the answers to the questions, as well as the relevant questions, are constantly changing as oil analysts track a constantly moving target. Methodologies developed to study past market changes and suggest future market directions have evolved as modelers identify and begin to study the extremely complicated set of parameters that determine oil prices. But foremost, oil forecasters increasingly recognize the vulnerability of their trade to conceptual and empirical uncertainties. Users of forecasts are increasingly warned that projections reflect specific assumptions about the structure of the oil market and the key· market parameters that are dictated by the use of a particular methodology. If history has not necessarily made forecasters of oil prices more accurate, it has made them wiser. C. [ [ D [ [ [ c r L [ [ c 0 [ c r u [ L -. L 75 4 REFERENCES Adelman, Morris. 1982. "OPEC as a Cartel," In Griffin and Teese, pp. 37-63. . Alaska Power Authority. 1984. "Comments on the Federal Energy Regulatory Commission Draft Environmental Impact Statement of May 1984," Document No. 1774, Susltna Hydroelectric Project, August. American Petroleum Institute. 1984. Basic Petrole!ID Data Book. vol. 4, no. 1, Washington, D.C., January. Aperjts, Dimitri. 1982 The or Pol Icy d E .I Market In the 1980s: OPEC Ot I an conorntc Deyelogment B I Massachusetts.· a linger, Cambridge, Betder, Perry. 1981. Energy Mode I I ng Call foro I a. "A Comparison of the EMF-6 Models," Mlmeo, Forum, Stanford University, Stanford, Soh I, Doug I as R. and W II IJ an B. Quandt. 1984. Energy SecurIty [ n the 1980s: Economic and Pol tttcal Perspecttyes, The Brookings Institution, Washington, D.C. Bohl, Douglas R. and Milton Russell. 1978. Limiting Oil lmoorts; An Economic History and Analysts, Johns Hopkins University Press for Resources for the Future, Baltimore. Boh!, Douglas R. and w. David Montgomery. 1982. 011 Prices. Energy Security. and Import Pol ley, Johns Hopkins University Press for Resources for the Future, Baltimore. B I I tzer, C. , A. Meeraus, and A. Stout jesd I j k. 1 975. "A DynamIc Model of OPEC Trade and Product! on," Journal of Deyel opment Econan!cs, vol. 2, no. 4, pp. 319-335. Braden, David. 1981. Simulation Models Forum, EMF-6.11. January. "A Conceptual Framework for Understanding of the Wor I d 0 I I Market," Energy Model I ng Stanford University, Stanford, Cal lfornla, Central Intel I lgence Agency. 1984. International Energy Statistical Reyfew, Washington, D.C., September 25. Chase Manhattan Bank. 1984. Economic Obseryer, vol. 4, no. 5, New York, September/October. Cremer, Jacques and Djavad Salehl-lsfahanl. Pricing In the World Oi I Market: How mlmeo, Unlversl ty of Pennsy Ivan! a. 1980. "Competitive Important Is OPEC?" [ [ ~ c [ [ [ E c L [ [ c E [ c rc L [j L l 76 4 Cremer, J. and M. Weitzman. 1976. "OPEC and the Monopoly Price of 011," European Economic Review, vol. 8. no. 1, pp. 155-164. Curlee, T. Randall. 1983. "The Impacts of Ott Disturbances: Lessons from Experience," ORNL/TM-8492, Oak Ridge National Laboratory, Oak Ridge, Tennessee, January. Curlee, T. Randal I. 1984. Strategic Stockpile Acquisitions and the Current World Oil Market," Energy Pol Icy, vol. 12, no. 2, June, pp. 216-219. Curlee, T. Randal I. forthcoming. "Monitoring the Potential for Future 01 I Supply Crises: The Lack of Information," The Joyrnal of Energy and Deyel opment. Daly, George, James M. Griffin, and "Recent OTt Price Escalations: Stability," Jn Griffin and Teece. Henry B. Steele. lmpl !cations for 1982. OPEC Data Resources, Inc. 1983. Evolution In the Environment," prepared Development Office. "The Imp! !cations for Navy PI annlng of Domestic and International Energy for the Navy Energy Research and Data Resources, Inc. 1 984. "World 01 I Out I ook," Autumn. Deece, David A. and· JosephS. Nye. 1981. Energy and Security. Bat I Inger, Cambridge, Massachusetts. Devarajan, Shantayanan, R. Glenn Hubbard, and Robert Weiner. 1983. "Report to the Department of Energy of the . Stockp I I e Coordination Project Harvard Energy Security Program," Department of Energy, Washington, D.C., October 1 [DOE/CS/10047-T8]. Eckbo, P. L. 1976. The Future of World 011. Ballinger, Cambridge, Massachusetts. Energy Information Administration. 1981. "World 011 Market Outlook: Recent History and Forecasts of World 01 I Prices," Department of Energy, Washington, D.C. [DOE/EIA-0307]. Energy Information Administration. 1983. 1982 Annyal Energy Outlook. Department of Energy, Washington, D.C., April [DOE/EIA-Q383C82>]. Energy Information Administration. 1983. Mgdel Documentation Report, Department D.C., June [DOE/EIA-0412]. 011 Market Simulation of Energy, Washington, Energy Information Administration. 1984. Annual Energy Outlook ~ Deeartment of Energy, Washington, D.C., May [DOE/EIA- 0383(83)]. c. [ [ D [ [ [ c r L [ [ c c [ c r~ L c L [ 77 • Energy lnforatlon Administration. 1984. "Oil Market Forecast-- DRAFT." Department of Energy, Washington, D.C., November. Energy Model lng Forum. 1982. World Oil, EMF-6, Stanford Unlversliy. -Stanford,· Cal rtornl a. February. Ezzat!, AI I. 1976. "Future OPEC Price and ProductiOn Strategfes as Affected by lts Capacfty to Absorb 011 Revenues," European Econanlc Revlew, vol. 5, no. 3, July, pp. 190-194. Foster, W. G., L. Burton, and F. 0. Hanpeter. 1983. "Survey of Of I and Gas Supply-Demand Forecasts," Electric Power Research lnstltute, March [EPRI-EA-2994]. Gately, Dermot. 1983. "OPEC: Retrospectlve and Prospects: 1973- 1990," Eyropean Econgmlc Review, vol. 21, no. 3, May, pp. 313- 332. Gately, Dermot. 1984. "A Ten Year Retrospectlve: OPEC and the World 0{1 Market," Journal of Econgmlc Literature. vol. 22, no. 3, September, pp. 110Q-1114. Gately, Dermot, John F. Kyle,_ and Dletrlch Flscher. 1977. "Strategies tor OPEC's Pricing Decisions," Eyropean Econgmlc Review, vol. 10, pp. 209=230. GrlffiD, James M. and David J. Tease, eds. 1982~ OFEC Behaylor and World Oil Prices, George AI len and Irwin, London. Hnylllcza, Estevan and Robert· S. Pfndyck. 1976. "Prlcfng Pol fetes for a Two-Par-t Exhaustible Resource Cartel," European Econanfc Rev lew. vol. 8, pp. 1 55-164. Horwfch, George and David Leo Weimer. 1984. 011 Price Shocks. Market Response. and eonttngency Planning. American Enterprise Institute, Washington, D.C. Hoteiiing, Harold. 1931. "The Econan!cs of Exhaustible Resources," Journal of Pol !tical Econaey, vel. 39, April, pp. 137-175. Huntington, Hillard G. 1984. "Real 0!1 Prices fran 1980 to 1982," The Energy Journal. vel. 5, no. 3, July, pp. 119-131. Johany, All. 1 980. The r-tJth of the OPEC Cartel, Wiley and Sons, New York. Kalymon, B. A. 1975. "Econanlc Incentives In OPEC 011 Pricing," Journal of Oeyelocment Econgmlcs. vel. 2, no. 4, pp. 337-362. Kask, Don E., Edward C. Fox, and "World Econanrc Recovery Could Thanos J. WII banks. 1983. Determine When the Next Energy c [ [ E [ [ [ c c L [ L c E [ c r L l L [ 78 Crtsts Wtll Occur," Ot! and Gas Journal. vol. 81, no. 51. December 19, pp. 82-91. Landts. Robert C. and Michael W. Klass. t 980. OPEC: Pol Icy Imp! !cottons tor the United States, Praeger, New York. Levy, W. J. 1974. "lmpl Jcattons of World Ot I Austerity," mtmeo, 30 Rockefeller Plaza, New York. Ltchtblau, John H. 1982. "The Limitation to OPEC's Pricing Pol tcy," to Grtffln and Teece, pp. 131-144. MacAvoy, Paul W. 1982. Crude Oil Prices: As Determined by O?EC and Market Fundamentals. Ballinger, Cambridge, Massachusetts. Manne, Alan S. 1984. "International Energy Workshop: A Progress Report," presented at the Meetings of the International Assoctatton of Energy Economists, San Francisco, November 5-7. Manne, Alan S. and Leo Schrattenholzer. 1984. "International Energy Workshop: A Summary of the t 983 Poll Responses," IbJt Energy Joyrna!. vol. 5, no. 1, January, pp. 45-64. Marshal Ia, Robert A. 1982. "World OJI Model," Dectston Focus Incorporated, Los Altos, Cal tfornta, October. Marshall a, Robert A. and Dale M. Nesbttt. 1984. "An Eval uatton of World 011 Pricing Strategies Ustng the Dectston Focus Incorporated World 011 Model," working paper, Decision Focus Incorporated, Los Altos, Cal lfornla. Mead, W. J. 1 g"/9. "The Performance of Regulations," American Econantc Review. pp. 352-356. Government Industry vol. 69, no. 2, May, Moran, Theodore. 1982. "Modeling OPECBehavlor: Econanlcand Pol Itlcal Alternatives," In Griffin and Teece, pp. 94-130. Netschert, Bruce C. 1983. "The Outlook for Future 011 Prices," Energy Exploration and Exploitation, vol. 2, no. 2, pp. 133- 1 42. Office of Policy and Evaluation. 1980. "Reducing 011 Vul nerab I I I fy: Energy Po I Icy for the 1 980s," Department of Energy, Washington, D.C., November 10. Plndyck, Robert S. 1978. "Gains to Cartelization of Exhaustible Resources," and Statistics, vol. 60, May, pp. 238-251. Producers Rey lew of £rem the Econanlcs Plndyck, Robert S. 1979. "Some Long-Term Problems In OPEC 011 Pricing," The Journal of Energy and Deyelo~eot, vol. 4, no. 1, pp. 259-272. c [ [ c [ [ [ c r 6 [ [ c [ E c r u [ L [ 79 Plndy~, Robert S. 1982. "OPEC Otl Prlctng, and the lmpl !cations tor Consumers and Producers," tn Grlfftn and Teece, pp. 175- 185. Rot.masset, J., 9· Isaak,· and F. Fesharakt. 1983. "011 Prices Wtthout OPEC: A Walk on the Supply Stde," Energy Economics. vol. 5, no. 3, July, pp. 164-170. Sal ant, Stephen W. 1982. tmpertect Oompetltton In the World 01 I Market. Lexington Books, Lexington, Massachusetts. Satty, Than as L. and A. H. Ghol amnezhad. 1981. "01 I Pr fees: 1985 and 1990," Energy Systems and Polley, vol. 5, no. 4, pp. 303- 318. Satty, Thanas L. 1980. The Analytl c HIerarchy Process, McGraw- Hi II I nternatl onal, New York. Singer, S. Fred. 1982. "The World Energy Outlook," In Energy Model log IV: Planning for Energy Disruptions. Institute of Gas Technology, Chicago, pp. 567-575. - Stnger, S. Fred. 1983. "World Energy Outlook," In ~ HYdrocarbon Markets: Current Statys. Pro!ected Prospects. and Futyre Trends. M. S. W!onczed, ed., Pergamon Press, Oxford, England. Sweeney, J. L. 1983 •. "Modeling the World 0!1 System," In Use of Slmylatlon Models In Energy Planning: Proceedings of the International Qonterence on the Use of Simulation Models !n Energy Planning, Energy Systems Group, Rtsoe National Laboratory, Rosk t I de, Perm ark. Teece, Dav!d J. 1982. "OPEC Behav!or: An Alternative View," to Grtftln and Teece, pp. 64-93. Texaco. 1983. "Free World Energy Survey: Historical Overview and Long-Term Forecast," Finance and Econanlcs Department, December. Vernon, Raymond, ed. 1976. Jbe OJ I Crisis. W. W. Norton and Company, New York. Weyant, John P. 1983. "The Energy Crtsts Is Over ... Agatn," Cbal lenge, vol. 26, no. 4, September/October, pp. 12-17. Weyant, John P. 1984. "The Conttnutng Threat of 011 Supply Interruptions," Journal of Polley Analysts and Management. vol. 3, no. 3, pp. 393-405. c [ [ E [ [ [ D r 6 [ [ D E E c r LJ L L l 4 ORNL/TM-9521 INTERNAL DISTRIBUTION 1. D. J. Bjornstad 20. H. C. Mei 2. V. 0. Baxter 21. R. 0. Perlack 3. c. R. Boston 22. c. H. Petrich 4. S. Cantor 23. L. w. Rickert 5. F. C. Chen 24. G. Samuels 6. J. E. Christian 25. R. B. Shelton 7. T. R. Curlee 26. R. A. Stevens 8. G. A. Dailey 27. G. G. Stevenson 9. R. M. Davis. 28. B. E. Tonn 10. w. Fulkerson 29. o • P • Vogt 11. D. L. Greene 12. G. R. Hadder 13. 0. M. Hamblin 14. c. E. Hammins 15. L. J. Hill 16. E. L. Hillsman 17. R. B. Honea 18. R. Lee 19. W. McConnell 30. G. P. Zimmerman 31-53. Energy and Technology Economics Group 54. Central Research Library 55. Document Reference Section 56. Laboratory Records 57. Laboratory Records (RC) 58. ORNL Patent Office EXTERNAL DISTRIBUTION 59. Office of Assistant Manager, Energy Research and Development, DOE-ORO, Oak Ridge, Tennessee 37830 60. s. Malcolm Gillis, Professor, Economics and Public Policy, Department of Economics, Duke University, Durham, North Carolina 27706 61. Fritz R. Kalhammer, Vice President, Electric Power Research Institute, P. 0. Box 10412, Palo Alto, California 94303 62. Todd R. LaPorte, Professor Political Science, Institute of Government Studies, University of California, 109 Moses Hall, Berkeley, California 94720 63. William H. Williams. Division Manager, AT&T Information Systems, Building 83, Room 1823, 100 Southgate Parkway, Morristown, New Jersey 07960 64-88. Technical Information Center, DOE-ORO, Building 1916 T-1, Oak Ridge, Tennessee 37830