|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
Abstract Conventional oil and gas price forecasts typically include pessimistic, most-likely, and optimistic cases in an attempt to quantify economic uncertainty. An analysis of forecasts presented by industry and governmental organizations illustrates that conventional forecasting methods typically underestimate significantly the full range of uncertainty in oil and gas price forecasts. Economic indicators calculated using such forecasts will not reliably quantify investment risk. In this investigation we compared and contrasted several recently developed methods for quantifying upstream petroleum investment risk due to uncertainty in future prices. We analyzed five forecasting techniques - conventional, bootstrap, Inverted Hockey Stick (IHS), historical, and Sequential Gaussian Simulation (SGS). These techniques were applied to three synthetic projects and 23 industry projects to examine the uncertainty associated with economic indicators such as net present value, investment efficiency, and internal rate of return. Across all 26 projects, the conventional forecasts predicted a narrower range of economic indicator values than the four alternative methods, indicating that conventional methods routinely underestimate uncertainty. All four alternative forecasting techniques can provide operators with more reliable quantification of the uncertainty inherent in their investment decisions than provided by conventional methods currently in widespread use. The four alternative methods have unique strengths and weaknesses that may affect their applicability in particular situations. The SGS methods is the most rigorous and accurate method; however, it is also the most difficult to apply. The IHS method serves as a reasonable approximation, and can be easily incorporated into existing procedures and software. Introduction Investments in the petroleum industry are made under significant uncertainty. According to Capen, uncertainty is underestimated on an almost routine basis. Stermole and Stermole note that uncertainty will be a factor "no matter how comprehensive or sophisticated an investment evaluation may be." Experts have stated that oil and gas producing assets are subject to three classes of uncertainty: technical, political, and economic. Economic uncertainty affects investments within the petroleum industry at least as much as its technical counterpart. Unlike technical uncertainty, which should decrease with production of a reservoir, economic uncertainty does not decrease over the life of a petroleum reservoir. Future oil and gas prices represent a substantial source of economic uncertainty for operators considering exploration and development opportunities. Wiggins claims that price projections are as important as reserves determinations and production forecasts when evaluating hydrocarbon-producing properties. Campbell et al.  affirmed that errors in project valuations are more attributable to price forecasts than any other component. Although we cannot eliminate uncertainty from investment evaluations, we can better quantify the uncertainty by accurately predicting the volatility in future oil and gas prices. Reliably quantifying economic uncertainty will enable operators to make better decisions and allocate their capital with increased efficiency. Price projections within the petroleum industry are often comprised of pessimistic, most-likely, and high cases in an attempt to quantify uncertainty. Typically, these forecasts initially decline or remain flat for a period of time before increasing monotonically. Such price projections are referred to as "hockey stick" forecasts. The California Energy Commission (CEC) published a natural gas price forecast beginning in 1997 that clearly illustrates the characteristic "hockey stick" shape commonly exhibited by conventional price forecasts (Fig. 1). A segment of the CEC natural gas price forecast beginning in 2002 is shown in Fig. 2 along with actual gas price data realized by the market. The CEC forecast clearly underestimates the true range of product price uncertainty. A considerable portion of the actual gas price data falls outside of the high and low extremes presented in the forecast.
Society of Petroleum Engineers 6200 North Central Expressway Dallas, Texas 75206
THIS PAPER IS SUBJECT TO CORRECTION
American Institute of Mining, Metallurgical, and Petroleum Engineers, Inc.
A practical general approach to forecasting foreign oil prices to the year 2000 is described, which includes both non-market and market influences and their relative importance. Using the approach, actual prices are projected, and an estimate of possible range of variation from the forecast is given. Also, forecasts of other are discussed.
In essence, foreign oil prices, expressed in constant 1975 U.S. dollars per barrel, are projected to decline somewhat by about 1985, projected to decline somewhat by about 1985, and then to rise sharply as the end of the century approaches. The possible percentage variations from the projections are large early in the period, and, surprisingly, are expected to decline later in the period.
When I say long range forecasting, what I mean is to the year 2000, or about a 25-year forecast. I think I hardly need to mention the hazards and uncertainties in long range forecasting, especially today and especially in energy. About the only type of forecasting I consider more hazardous is very short range forecasting, because in that kind of forecasting people still remember what you said when it people still remember what you said when it doesn't happen.
In the next quarter century, I expect foreign oil prices to be determined by a number of different major influences which vary substantially in importance with time. Putting this thought another way, I don't think that there is any single forecasting approach which will work over such a long period of time.
For lack of a better term, I have chosen to classify future influences on foreign oil price into two major groups. First, there are price into two major groups. First, there are what I would like to call conventional market factors, which are the usual supply and demand factors which influence price in a competitive, unconstrained market. These include, for example, such factors as the downward influence on demand of an increase in price. second, there are what I would like to call non-market factors. These include all the other influences on price, such as political, military and financial influences, as well as competitive restraints such as cartels. Some economists might prefer to use the term exogenous factors instead of non-market factors. In general terms, at present and for about the next ten or so years, I see these latter, non-market influences dominating foreign oil price. Then, later in the century, I see the influence of conventional market factors increasing, but with non-market influences still substantial.
The great and continuing importance of non-market factors, which are virtually impossible to quantify, has, in my view, made precise modeling and computer techniques for forecasting of foreign oil prices impractical. I am not trying to belittle the usefulness of quantitative techniques for evaluating conventional market influences, or for evaluating some specific non-market influences.
Earlier explorations of how to predict oil prices led, in 1978, to my being named project director of a contract between USC and the OPEC Secretariat in Vienna to develop a world energy model. OPEC wanted a tool to help devise rational oil production and pricing policies. Late in 1981, we installed the model on OPEC's computers in Vienna. It is unknown how effectively OPEC was able to use the model, but it did foretell the shape of recent events in world oil markets. In Jan. 1981, we gave our sponsors the results of a study of what would happen if OPEC raised its production rate back to 20 million barrels per day, a study they requested after their production rate fell to around 18 million B/D. The model predicted that if OPEC's rate were to be raised as proposed, oil price would fall $20/bbl within a year. Although not spot on, the prediction is certainly indicative of what was to come. (For this study all price elasticities were arbitrarily increased by 10 percent from their base levels, the medians of estimates reported in the literature. In hindsight, this adjustment was conservative. The past decade's price-induced, reduced growth rate in energy consumption, so-called conservation, corresponds to even larger values of price elasticity.) There is no indication that OPEC paid attention to this or other model predictions, they were still whistling their favorite high-price tune when the market forced their first official price cut late in 1982.
This paper presents some rules of thumb that my experience suggests may help to understand the evolution of oil prices. The intent is to identify factors to consider in deciding on the credibility of an oil price forecast (yours or another's), rather than to prophesize future oil price. The presentation is professorial in style; a series of questions penetrating different aspects of the dynamics of oil prices are asked, and answers to these queries are advanced. This interrogatory is targeted to broaden the reader's perspective on what makes the oil market tick. We also examine how individuals have responded in the past to signals from the marketplace. This examination's purpose is to call to attention the impact of emotions on the subjective weightings applied in reaching an economic decision. In the oil patch expectations about highly uncertain outcomes, both economic (mostly prices and taxes) and physical, exert a major influence on investment decisions. Being aware of how emotions can slant these expectations is a key consideration for a decision maker.
To establish a backdrop for our discussions the first question asked is, "What is the consensus, if any, of recently published prognostications of future oil prices?" Sorting through the stream of predictions appearing in recent issues of Petroleum Intelligence Weekly, Oil and Gas Journal, et al., one finds these major themes:
ISSUES IN PETROLEUM SUPPLY AND DEMAND-VIEWS OF THE YEAR 2000 Thomas G. Burns, Manager, Economics, Chevron Corporation, San Francisco, California, USA. Abstract. What are the economic and political forces that will influence both the supply of and the demand for petroleum? Production techniques, exploration potential, and the role of enhanced recovery will influence the supply side. On the demand side of the equation, conservation, substitution and efficiency improvements will continue to play a major role. Understanding the basic limits imposed by economics and technology helps to circumscribe the arena in which political power can be exercised to influence oil markets. The integration of these factors into a coherent picture helps to define a number of plausible views of the developing petroleum supply/demand/price picture. Planners must evaluate all of the possible alternatives in order to test the ability of their plans to cope with the acknowledged uncertainties of the marketplace. Résumé. Quelles sont les forces économiques et politiques qui vont influencer à la fois l'offre et la demande de pétrole? Les techniques de production, le potentiel d'exploration et le rôle de la récupération assistée vont influencer l'offre. Quant à la demande, les améliorations concernant les économies d'énergie, le remplacement d'énergie et l'efficacité continueront à jouer un rôle extrêmement important. La connaissance des limites fondamentales imposées par la rentabilité et la technologie aide à circonscrire le domaine dans lequel le pouvoir politique pourra être utilisé pour influencer les marchés du pétrole. L`intégration de ces facteurs dans un tout cohérent aide à définir plusieurs vues probables de l'ensemble de l'évolution de l'offre, de la demande et des prix du pétrole. Les responsables devront évaluer toutes les solutions possibles afin d'examiner dans quelle mesure leurs projets permettent de faire face aux incertitudes reconnues du marché.
From its earliest days, the oil industry has been subject to the extreme pressures of the marketplace. In many respects, stability and continuity have been the exception rather than the rule. Brief periods of stability have more often than not been punctuated by periods of extreme volatility and instability in oil markets. There are several factors inherent in the oil business which tend to create this phenomenon. Probably the most significant contributing factor is the underlying cost structure of the oil industry. Before trying to describe a set of possible views of the petroleum industry in the year 2000, it is necessary first to analyse some of the fundamental issues shap- ing the industry-past, present and future. Evaluation of the responses to a series of questions-some implicit in the `conventional wisdom' of the industry and some raised by `contrarians'-may help to clarify these issues. 1. ARE OIL PRICES INHEREN