NOTE NUMBER 275 P U B L I C P O L I C Y F O R T H E privatesector OCTOBER 2004 Crude Oil Prices Robert Bacon and Predicting Price Differentials Based on Quality Silvana Tordo M a ny d eve l o p i n g c o u n t r i e s a re b e c o m i n g o i l ex p o r t e r s , p ro d u c i n g Robert Bacon c r u d e o i l s t h at o f t e n d i f f e r m a r ke d l y i n q u a l i t y f ro m t h o s e p r i n c i p a l l y (rbacon@worldbank.org) is the manager, and t r a d e d . G ove r n m e n t s m u s t p re d i c t t h e p r i c e s o f s u c h c r u d e s t o Silvana Tordo T H E W O R L D B A N K G R O U P PRIVATE SECTOR DEVELOPMENT VICE PRESIDENCY f o re c a s t reve n u e a n d eva l u at e t h e f a i r n e s s o f t h e p r i c e t h ey re c e i ve (stordo@worldbank.org) a senior energy economist, f ro m c o m p a n i e s s e l l i n g o n t h e i r b e h a l f . O i l c o m p a n i e s a n d i n d u s t r y in the World Bank’s Oil c o n s u l t a n t s h ave m o d e l s f o r a n a l y z i n g p r i c e d i f f e re n t i a l s w i t h we l l - and Gas Policy Division. Anna-Maria Kaneff of the k n ow n “ m a r ke r ” c r u d e s , b u t t h e s e m o d e l s h ave n o t b e e n w i d e l y International Finance k n ow n o r a d a p t e d t o a c c o u n t f o r i n c re a s i n g l y i m p o r t a n t q u a l i t y Corporation’s Oil, Gas, c h a r a c t e r i s t i c s s u c h a s a c i d i t y. T h i s N o t e ex p l a i n s a m e t h o d o l og y f o r Chemicals, and Mining Credit Review and p r i c e a n a l y s i s a n d a n ew ex t e n s i o n f o r i n c o r p o r at i n g a c i d i t y, w h i c h Portfolio Division assisted c a n h ave a b i g e f f e c t o n t h e p r i c e d i f f e re n t i a l . with data collection. The world produces and trades more than 160 all these cases governments need to understand varieties of crude oil, which range widely in the revenue implications of new production price: while the United Kingdom’s Brent Blend streams, because these can account for a large averaged US$43.04 in August 2004, Syrian share of their budget. Long-term price forecasts Heavy averaged US$29.97. Such large differen- are regularly published for key crudes—such as tials show why a single price cannot serve as a Brent Blend, West Texas Intermediate, or forecast for all crudes.1 Dubai—but no forecasts are published for other In recent years many new producer countries crudes. Instead, forecasts for a “marker” crude are have started to export substantial amounts of adjusted by a discount (or premium) specific to crude oil, including Azerbaijan, Chad, Equatorial each crude that is based on various aspects of Guinea, Kazakhstan, and Sudan. Several others quality. are promoting petroleum exploration, with good Under production sharing contracts the gov- prospects of finding commercially viable amounts ernment often relies on the oil company to sell of oil. And existing producers are opening new its share. Similarly, under royalty agreements fields where the quality of oil can differ greatly the concessionaire usually sells the oil and trans- from that of fields already under production. In fers the royalties, linked to the price of crude, to C R U D E O I L P R I C E S PREDICTING PRICE DIFFERENTIALS BASED ON QUALITY the government. Since the price governments index that includes various types of acid. Some of receive may differ, sometimes substantially, these acids pose no particular problems in the from the well-publicized marker crude prices, refinery process. But above a certain limit acidity they need to understand the factors that influ- has a corrosive effect on refineries. Blending low- ence the discount. For new producer countries, TAN with high-TAN crude can deal with this where the quality of crude may differ substan- problem, but it increases logistical costs. New tially from that of the marker crudes, this under- refineries constructed using special materials can standing is critical for projecting revenue and tolerate higher acidity but are few in number. establishing more trusting relationships with oil Thus crudes with a high TAN (greater than 2 companies. around 0.5), because they limit the options for refining, are likely to command a discount. Defining quality In recent years high-TAN crude has Crude oils differ from one another in a large accounted for a growing share of the global oil number of chemical and physical properties, supply: an industry study reports that crude with many of which play an important part in their a TAN greater than 1.0 increased from 7.5 per- refining and subsequent sale as petroleum prod- cent of the global supply in 1998 to around 9.5 ucts. Until now statistical analysis of price dif- percent today.2 For West African crude the pic- ferentials has focused on two main properties: ture is even more dramatic: in 2001 only 5 per- the specific gravity (lightness) measured in cent had a TAN greater than 1.0, but by 2006 degrees API (a scale devised by the American this share is expected to rise to 13 percent. If a Petroleum Institute) and the percentage of sul- high TAN means a substantial discount, this fur content by weight. forecast has important implications for devel- Lighter crudes (with higher API) produce a oping countries depending on oil for a large larger number of lighter products, such as gaso- share of government revenue. Moreover, if the line, which have higher resale value. So other share of high-TAN crude grows more rapidly qualities being equal, lighter crudes are than the refineries able to accommodate it are expected to sell at a premium over heavier constructed, the discount can be expected to crudes. By extension, if the prices of all petro- increase. leum products rise by the same percentage amount, the absolute price differential between Relating quality to discounts a heavy crude and a light crude (the discount) Oil companies and analysts have developed can be expected to grow. models for predicting discounts, some of which A high sulfur content has an adverse effect on are commercially available as off-the-shelf prod- the value of crudes, because it leads to higher ucts. But these models have centered on API operating costs for refineries due to special pro- and sulfur content. The growing importance of cessing and maintenance requirements. In addi- high-TAN crudes, especially from developing tion, in many countries new legislation mandates countries, calls for extending these models. Two lower sulfur content for gasoline and diesel. So approaches have been used. The first is based on a high-sulfur (sour) crude is expected to sell at a a simple pairwise comparison of crudes identi- discount relative to a low-sulfur (sweet) crude of cal in all major aspects of quality except the one the same API. And an increase in the share of being investigated. The second is based on a high-sulfur crude in the world market, or a rela- regression model that can simultaneously incor- tive increase in the demand for low-sulfur prod- porate all quality differences. ucts, is expected to result in larger discounts for If crude oils differ in only one dimension of high-sulfur crudes. quality, pairwise comparisons can be used to Another important property is acidity. The relate the price differential to that difference in recent emergence of West African and other pro- quality. But any single comparison would be ducers has led to an increase in the supply and affected by factors relating to the period or to number of crudes with high acidity as measured nonquality characteristics of the crudes being by the total acid number (TAN), an aggregate compared. As a result, a series of comparisons for several different crudes or for several differ- The model as presented above is static. Thus ent periods will not yield identical results, so that it would predict that the set of price differentials the results need to be averaged over a group of between all crudes would stay constant as long as crudes. This approach can be extended to com- the qualities of the crudes remained constant. In parisons of crudes differing in more than one fact, a plot of the average price difference dimension of quality by utilizing a series of pair- between crudes, period by period, shows sub- wise comparisons. stantial movement, roughly matching the move- A regression model overcomes these difficul- ment of the Brent price, with a greater spread of ties by simultaneously incorporating differences differentials at higher prices. To allow for the 3 in all important qualities in relation to the price hypothesis that a given difference in API has a differential. Assuming that the effects of a one- greater effect at higher prices, the API differen- unit difference in quality are the same regard- tial term is modified by scaling it by the current less of the crudes being considered, and that the Brent price. Another extension to the model effects of differences in several measures of introduces a lagged value of the price differen- quality are additive, the discount effect can be tial into the equation, recognizing that price dif- formulated as follows: ferentials may not fully adjust within a month (the data period used) to a change in the aver- ∆ price = a + Σi bi (∆ qualityi ) age Brent price within that period, but that some adjustment may spill over into the next month. where ∆ price is the difference in price between a given crude and some marker crude, ∆ quali- Estimating the model tyi is the difference in quality i between that Estimating the model required identifying the crude and the marker crude, Σ represents the crudes for which data on both quality specifica- sum of the qualities (three in the case consid- tions and spot prices were publicly available for ered below), and a and bi are parameters to be a number of months. Data on 56 crudes were col- established by regression analysis. lected for the period January 2003–June 2004, In this model the parameter b measures the with data on quality taken from several Web sites effect on the price differential of a one-unit and data on monthly average prices (fob) taken increase in the difference in each dimension of from the Platts Oilgram Price Report. quality between the crude in question and the Brent Blend was taken as the reference crude, marker crude. since it is among the most common markers for Although a regression model will not fully commercial deals. Brent is a high-quality light explain each difference in price in terms of dif- crude (API of 38.3) with low sulfur (0.37 percent) ferences in quality, evidence for a large number and a very low TAN (0.07), so that most other of crudes can identify effects common to all. crudes would be expected to sell at a discount rel- Moreover, since the effects (other than those ative to it. During the period analyzed the price of relating to the general oil price level) are not Brent ranged between US$25 and US$38 a barrel. expected to change rapidly, the model should The model that was estimated includes sev- be applicable to a succession of intervals and the eral refinements to the simple framework values of the coefficients should be stable. So described above: data for a number of crudes and for a number ■ The difference in API is multiplied by the of periods can be used in the same regression current Brent price because the absolute dif- (utilizing a pooled cross-section time-series ference in the lightness of the crude is approach). The quality characteristics of each expected to lead to a larger absolute price dif- crude are assumed to remain stable throughout ferential when the general oil price is higher. each period, since there is no published evi- ■ The difference in TAN is included only for dence on changes over time. But if the period crudes whose TAN is above 0.5, allowing for used is too long, the physical characteristics of the assumption that below a certain thresh- crude oils, especially blends, may change old refiners need to take no special action. enough to introduce inaccuracies. ■ Freight rates between Europe and the region C R U D E O I L P R I C E S PREDICTING PRICE DIFFERENTIALS BASED ON QUALITY Box Checking the price differential for Syrian Heavy crude in June 2004 1 The case of Syrian Heavy illustrates how the model can be used to estimate what a price would have been given the model’s assumptions. Syrian Heavy has low API (24.1), a high sulfur content (3.9 percent) but a TAN (0.17) below the threshold at which the model takes it into account. viewpoint Price of Brent Blend in June 2004 = US$35.04 Discount in May 2004 = –US$5.93 Effect of lagged discount = 0.67 (5.93) = –US$3.98 is an open forum to Effect of lower API = 0.00154 x 35.04 (24.1 – 38.3) = –US$0.77 encourage dissemination of Effect of higher sulfur = 0.27 (3.9 – 0.37) = –US$0.95 public policy innovations for Effect of higher TAN = 0 (0.17 – 0.07) = –US$0.00 private sector–led and Effect of transport differentials = –US$0.28 market-based solutions for Total predicted discount relative to Brent = –US$5.98 development. The views Actual discount relative to Brent = –US$6.24 published are those of the Predicted Syrian Heavy price at actual Brent price = US$29.06 authors and should not be Actual Syrian Heavy price = US$28.80 attributed to the World Bank or any other affiliated The strong coherence between the prediction and the actual data confirms, within the margin of error of such a regression organizations. Nor do any of model and taking quality characteristics into account, that the price of Syrian Heavy conformed to those of other crudes. the conclusions represent official policy of the World Bank or of its Executive from which the other crudes originate are tions, to see whether there was anything unusual Directors or the countries introduced to allow for systematic differ- about the actual price received (box 1). they represent. ences due to transport costs and for varia- tions over time due to changes in transport Conclusion To order additional copies costs. This term also includes a constant A regression model relating price differentials contact Suzanne Smith, effect common to all crudes. to differences in quality characteristics can be a managing editor, Room F 4K-206, All variables are strongly significant and of valuable tool for analyzing the price of new The World Bank, the expected sign, and the overall squared cor- crudes whose quality differs greatly from that of 1818 H Street, NW, relation coefficient is 80 percent. The coeffi- the marker crude. This capability enables gov- Washington, DC 20433. cient of the lagged price differential is estimated ernments whose country produces such crudes at 0.67. The long-run effects of the variables, to better forecast future revenue streams and to Telephone: evaluated as if the Brent price settled at some verify that the prices received from companies 001 202 458 7281 constant level, are as follows: selling the oil on their behalf are reasonable. Fax: 001 202 522 3480 Per unit of sulfur –US$0.82 Email: Per unit of TAN –US$0.88 ssmith7@worldbank.org Per unit of API per dollar of Brent +US$0.0047 Notes This Note is taken from a larger study now under Produced by Grammarians, The interpretation of the API coefficient is preparation. Inc. particularly interesting. At a price of US$20 a 1. Conventionally the industry focuses on absolute barrel for Brent each degree of API is worth (dollar) differentials rather than relative (percentage) dif- Printed on recycled paper 9.4¢, while at a price of US$40 it is worth 18.8¢, ferentials. supporting the hypothesis that differentials 2. Anne Shafizadeh, Gregg McAteer, and John increase as crude prices rise. Sigmon, “High-Acid Crudes,” paper presented at Crude The model can be used to predict the price Oil Quality Group meeting, New Orleans, January 30, 2003 of future streams of crude of known quality [http://www.coqg.org/20030130special.asp]. under assumed Brent prices or to “backcast” by estimating what the price of a known crude would have been given the model’s assump- This Note is available online: http://rru.worldbank.org/PublicPolicyJournal