8.6. Using the Model to Generate Scenarios
Having described the conceptual building blocks and feedback loops of the oil producers' model we are now ready to simulate scenarios. The key to effective scenario modelling is to unfold several alternative futures (each an internally consistent story, but with a different plot and a different ending) in order to challenge conventional wisdom and to encourage users to think how they would act if this future were to unfold. It is important to remember that such simulated futures are not predictions of oil price, supply or demand, nor do they represent official company forecasts. Two simulations are presented below. Each covers a 25-year period from 1988 to 2012 using the version of the model without Russian oil, developed before the fall of the Soviet empire. The model, called 'Oil World 1988' can be found in the CD folder for Chapter 8. I have deliberately chosen to present these archive scenarios in order to show simulations exactly as they were originally presented to the project team. However, it is also instructive to repeat the simulations under the same scenario conditions in 'Oil World 1995'. This more up-to-date version of the model runs over a 25-year period from 1996 to 2020 and shows the profound effect of Russian oil reserves on the global oil market and long-term price dynamics.
8.6.1. Archive Scenario 1: 10-Year Supply Squeeze Followed by Supply Glut
Imagine that in 1988 OPEC had followed a policy in which quotas were set 10 per cent lower than the call on OPEC (the difference between total demand for crude oil and the independents' production). The policy is successfully enforced for 10 years until 1998, and then reversed. Hence, in the remaining 15 years of the simulation, from 1998 to 2012, quotas are set 10 per cent higher than the call. The result is a 10-year supply squeeze followed by a supply glut. Assume that OPEC maintains quota discipline throughout the squeeze and glut, so there is no cheating on quota by any member of OPEC. Also assume that non-price pressures on crude oil demand from the global economy and the environment are neutral so that, in the absence of price changes, demand would stay constant at 50 million barrels per day.
To run this scenario for yourself, open 'Oil World 1988'. The initial screen displays a blank time chart and the simulator controls for creating scenarios, as shown in Figure 8.18. Set 'cartel quota bias' to −0.1 by moving the slider to the left. This change brings about an OPEC supply squeeze. All the other controls should remain in their default positions. Press the 'Run' button and simulate for five years and then press the 'run' button once more to simulate for a further five years to 1998. At this point, reset the 'cartel quota bias' to +0.1 by moving the slider to the right in order to bring about an OPEC supply glut. Run the simulation out to 2012 and inspect the time charts. There are four pages available to view by clicking on the page tab symbol at the bottom left of the graph pad. The time charts are described below.
Figure 8.19 shows profiles of demand and production for the squeeze-then-glut scenario. Demand in the top chart begins at 50 million barrels per day in 1988 and falls gradually to about 45 million barrels per day during the decade of supply squeeze. As OPEC switches policy from squeeze to glut in 1998, demand begins to rise and maintains a steady growth to about 52 millions barrels per day by the end of the simulation in 2012. The lower half of the figure shows how demand is shared among the different producers. In 1988, the independents are producing 26 million barrels per day (line 1), the opportunists 17 million barrels per day (line 2), and the swing producer 7 million barrels per day (line 3). As the supply squeeze begins to bite OPEC's output gradually falls. By 1993 the opportunists have cut production by 5 million barrels to 12 million barrels per day, while the swing producer's output has fallen to 4 million barrels per day. Meanwhile the independents' production has risen to compensate.
Figure 8.18. Opening screen of Oil World 1988 showing simulator controls for scenarios
In 1994 and 1995 there is a temporary surge in production as the swing producer switches to punitive mode to prevent further loss of market share. At this point, the policy of supply squeeze is almost exhausted because the swing producer is unwilling to sacrifice further market share. The burden of any additional OPEC supply reductions is carried mainly by the opportunists. In 1997, there is another small burst of punitive production by the swing producer. Then in 1998, OPEC reverses policy, changing a supply squeeze into a supply glut. Gradually OPEC's output begins to rise. Opportunists' production increases from a low of 10 million barrels per day in 1998 to 21 million barrels per day in 2012. Meanwhile, the swing producer's output rises from 4 million barrels per day to 16 million barrels per day in 2012. It is interesting to note that OPEC's policy change from squeeze to glut has no immediate impact on independents' production. At the time of the policy change in 1998, the independents are producing 30 million barrels per day. Their combined output then rises to 31 million barrels per day by the year 2001. By 2003, 5 years after the policy change, independents' output is still 29 million barrels per day. Thereafter it declines steadily to 15 million barrels per day by the end of the simulation in 2012.
Figure 8.19. Demand and oil production in an archive scenario: OPEC squeeze then glut without Russian oil
Figure 8.20. Oil price and profitability of independents' capacity in an archive scenario: OPEC squeeze then glut without Russian oil
The impact of OPEC's policy on market oil price is shown in the top half of Figure 8.20. Oil price (line 1) begins at $15 per barrel in 1988. Broadly speaking, during the supply squeeze price rises to a peak of $30 per barrel in 1998, then falls during the supply glut back to about $15 per barrel. Within the rising and falling trend there is much interesting detail in the movement of price that arises from the interplay of feedback loops that balance supply and demand. The reader can view the state of supply and demand by comparing the trajectory of demand minus production (line 2) with the horizontal grid line denoting equilibrium. When line 2 is above the grid line, demand exceeds supply. When line 2 is below the grid line there is a supply surplus and supply exceeds demand.
From 1988 to 1990, price rises sharply to $20 per barrel as the supply squeeze takes hold. Demand exceeds supply as shown by line 2. Price then declines slightly between 1990 and 1991 as falling demand creates a temporary supply surplus (despite OPEC's policy). Then as the supply squeeze bites once more, price rises steadily to the next peak of $25 per barrel during 1994. Price dips sharply in 1995 as the swing producer switches into punitive mode and rapidly expands production, creating a supply surplus. After the pulse of punitive production OPEC sharply curtails production which causes price to rise sharply during 1996 to a peak of $33 per barrel. The oil price is maintained around $30 per barrel for a period of three years. The era of moderately high price ends during 1998 as the swing producer once more switches into punitive mode and price falls rapidly to $20 per barrel. Over the next two years, the price level recovers slightly to a peak of about $24 per barrel in 2001 before settling into a slow and gently fluctuating decline to its final value of $15 per barrel as OPEC's new policy of excess supply takes effect.
The lower half of Figure 8.20 shows the changing investment incentives for independent producers that helps explain the rise and fall of independents' production. The hurdle rate for upstream investment projects (line 1) is assumed to remain constant at 15 per cent (0.15). At the start of the simulation, when the oil price is $15 per barrel, the profitability of new capacity (line 2) is 50 per cent (0.5), reflecting the initial low cost of reserves. It is financially very attractive to add new capacity, but over the three-year interval to 1991, profitability falls sharply to equal the hurdle rate. This fall in profitability occurs despite the rise in oil price from $15 to $20 per barrel, because the development cost of new reserves is rising even faster (the independents are operating on the steep shoulder of the cost profile). As price rises to its second peak of $25 per barrel in 1994, the profitability of new capacity increases slightly above the hurdle rate. The price plateau of $30 per barrel between 1996 and 1998 creates an attractive era for investment in which the profitability of new capacity rises to 25 per cent, well above the hurdle rate. Then, as the swing producer generates a pulse of punitive production, and as OPEC changes policy, profitability falls quickly to only 10 per cent in the year 2000. The investment climate continues to deteriorate throughout the remainder of the simulation. By the year 2011, the profitability of upstream investment has fallen to zero, meaning that the price of oil is now equal to independents' development costs. In the final year of the simulation, profitability actually goes negative.
Changing profitability drives independents' capacity as shown in Figure 8.19, but only after a lag that represents the time to develop a new field. The lag effect is most obvious following the attractive investment era between 1996 and 1998.
Independents' capacity is static at 30 million barrels per day (actually declining slightly) during the high price plateau. Then as profitability declines sharply between 1998 and 2000, independents' capacity rises gradually to 32 million barrels per day as new fields, approved earlier, begin to come on stream.
8.6.2. Archive Scenario 2: Quota Busting in a Green World
Imagine a world in which there are environmental pressures to reduce oil consumption – a green world. In the absence of price changes, these pressures are assumed to be sufficient to cut consumption by 10 per cent per year. Imagine too that the opportunist producers in OPEC are inclined to cheat on quota. They deliberately plan to build capacity 10 per cent greater than quota, and secretly use the excess capacity to produce more than quota whenever the opportunity arises. Assume that OPEC adopts a neutral quota policy, setting quotas that exactly match its estimate of the market call on OPEC. To create these scenario conditions set the effect of 'global economy and environment on demand' to −0.1 and the 'opportunists' capacity bias' to +0.1. Also, make sure to reset the 'cartel quota bias' to its default value of zero. The other sliders should remain at their default values.
Figure 8.21 shows profiles of demand and production for this scenario of quota busting in a green world. In the top half of the figure, demand begins at 50 million barrels per day in 1988 and falls gradually to a final value of 44 million barrels per day by 2012. The decline is due to the assumption of a green world. Despite the ceteris paribus green assumption of a 10 per cent cut in consumption per year, demand falls by only 6 million barrels per day over 25 years due (as we will see) to the compensating effect of lower price resulting from conservation and quota busting.
In the bottom half of the figure, the opportunists' propensity to cheat pays off as they expand capacity and increase their bargaining strength in quota negotiations. Their production rises steadily from a starting value of 17 million barrels per day to a final value of 33 million barrels per day. Most of this expansion comes at the expense of the independent producers who see their output fall from a starting value of 26 million barrels per day to only 7 million barrels per day at the end of the simulation. The swing producer too is squeezed by the opportunists at the start of the simulation, as output falls from 7 million barrels per day in 1988 to 4 million barrels per day in 1994. However, beyond 1994, and for the rest of the simulation, the swing producer uses mild pulses of punitive production to maintain market share and hold production at 4 million barrels per day. The first pulse occurs in 1995. The next pulse of punitive production occurs in 2006 and again in 2011.
Figure 8.21. Demand and oil production in an archive scenario: Quota busting in a green world without Russian oil
The top half of Figure 8.22 shows market oil price (line 1). Price drops quickly from its starting value of $15 per barrel in 1988 to $11 per barrel in 1990 as demand falls in response to green conservation pressures. After this initial sharp drop, price falls very gradually during the rest of the simulation (with minor fluctuations caused by the swing producer's bursts of punitive production) to a final value of $8 per barrel by the end of the simulation. Needless to say, the sustained run of low oil price generates a bleak environment for independents' upstream investment. In the lower half of Figure 8.22, the profitability of new capacity quickly falls below the hurdle rate of 15 per cent. By 1991, profitability is at 11 per cent, and continues to fall throughout the simulation to a final value of about 1 per cent in 2012.
Figure 8.22. Oil price and profitability of independents' capacity in an archive scenario: Quota busting in a green world without Russian oil
With no financial incentive for new investment, the independents' capacity (lower half of Figure 8.21) gradually declines as existing fields are depleted. Quota busting in a green world leads to a rapid demise of the independent producers. They end the simulation with no new capacity in the pipeline, a gloomy investment climate, and a market share of only 16 per cent (down from 52 per cent at the start of the simulation, and still declining).
8.6.3. Contemporary Scenario: Asian Boom with Quota Busting, Cautious Upstream Investment and Russian Oil
The previous simulations were archive scenarios created before the fall of Communism in an era when the expectation among industry experts was that Russian oil would never be globally traded. Circumstances have changed. Hence, we end with a contemporary scenario running from 1996 to 2020 that includes Russian oil. The scenario assumes steady upward pressure on oil demand arising from a sustained Asian boom in the giant developing economies of China and India. In this climate of growth, opportunist producers within OPEC are assumed to engage in quota busting while independent producers take a cautious approach to upstream investment, approving development of only the most profitable new fields.
To run this scenario for yourself, open 'Oil World 1995'. The initial screen is identical to Oil World 1988 in Figure 8.18 except that the time chart spans the period 1996 to 2020. There is also one extra control called 'time to build trust in Russia' that determines the rate at which commercial confidence builds to develop Russian reserves. To mimic demand pressure from an Asian boom, move the slider 'effect of global economy and environment on demand' to the far right so it takes a value 0.1. This setting means that consumers collectively want to use more oil. To activate quota busting within the OPEC cartel move the slider 'opportunists' capacity bias' to the right so it takes a value of 0.1. Also set the 'cartel quota bias' to a value −0.05 to enact a deliberate policy of moderate undersupply by OPEC. Finally, to represent cautious upstream investment by independents reset 'capex optimism' to 0.6. This setting has the effect of authorising only the most profitable 60 per cent of the upstream development projects whose estimated financial return exceeds the hurdle rate. The remaining two sliders ('time to build trust in Russia' and 'oil price bias') can remain at their default settings.
Press the 'Run' button to see this contemporary scenario play out in five-year intervals resulting in the time charts shown in Figures 8.23, 8.24 and 8.25. Remember that any scenario shows time paths that are internally consistent with the assumptions about conditions in the industry. These time paths do not necessarily fit history or point-predict the future. In Figure 8.23, demand for oil starts at 68 million barrels per day in 1996 and over the decade to 2006 rises gradually to 71 million barrels per day. The increase in demand is modest, only 4 per cent over ten years despite steady upward pressure on demand from the assumed growth in Asian economies. The reason that demand does not grow more quickly is that oil price (shown in the bottom half of Figure 8.23) rises from $16 per barrel to almost $30 per barrel – practically doubling over the simulated decade. Such a large price increase evokes greater fuel efficiency in the global economy.
Figure 8.23. Demand and oil price in a contemporary scenario: Asian boom with quota busting, cautious upstream investment and Russian oil
Figure 8.24. Production profiles and profitability of independents' capacity in a contemporary scenario: Asian boom with quota busting, cautious upstream investment and Russian oil
In the remainder of the simulation, from 2006 to 2020, demand grows swiftly and is also far more volatile. In only two years, from 2008 to 2010, demand surges from 72 to 79 million barrels per day, a 10 per cent increase, as global economic growth combines with a dramatic drop in oil price (from $30 to $15 per barrel) to unleash vast additional consumption. As we will see later the price drop is caused by a temporary burst of punitive production by the swing producer that floods the market with oil. When punitive production ends the oil price rapidly rises to $35 per barrel by late 2010 leading to a slight reduction of demand, despite the continued pressure of economic growth. Thereafter, demand continues on an upward trajectory with intermittent upturns and downturns, reaching almost 90 million barrels per day by 2020. The ups and downs stem from volatility in oil price as the oil industry enters an era of OPEC-induced supply instability between 2008 and 2020.
The supply story can be seen in Figure 8.24. The top chart shows the production profiles of the three main producer groups – the independents (line 1), the opportunists (line 2) and the swing producer (line 3). In the interval between 1996 and 2002, independents' capacity (and therefore production) is steady at almost 40 million barrels per day. Over the same six years, opportunists' production increases gently from 18 to 20 million barrels per day as opportunists take advantage of demand pressure and stable oil price to engage in limited quota busting. Meanwhile, the swing producer's production fluctuates gently and then settles into a downward trend resulting from the deliberate policy of undersupply and the need to compensate for quota busting by opportunists. The overall picture in the early years of the scenario is restrained production by the industry as a whole that leads to the near doubling of oil price already seen in Figure 8.23. High price creates a favourable environment for commercial upstream investment as shown by the trajectory for the profitability of new capacity in the bottom half of Figure 8.24. Profitability (line 2) begins at 12 per cent (0.12), just below the hurdle rate (line 1) of 15 per cent (0.15) in 1996 and rises steadily to 43 per cent (0.43) by 2008. As a result, more and more upstream projects are approved and independents' capacity (line 1 in the top chart begins to rise), reaching 47 million barrels per day by 2008 and 56 million barrels per day by 2014. Independents take an increasingly large share of demand and this relative growth puts pressure on the OPEC cartel. The swing producer is increasingly marginalised and finds production falling to only six million barrels per day by 2008 despite growth in demand.
This unsatisfactory condition of chronic low output provokes the swing producer into punitive production starting in early 2008. The indicator for swing mode (line 3 in the bottom chart) moves from one to zero and swing producer production surges upward reaching 14 million barrels per day in 2009. As we saw previously, oil price plummets unleashing latent growth in demand. Meanwhile, the opportunists cease quota busting and their production takes a temporary dip between 2009 and 2010. By 2010, the swing producer ceases punitive production having regained market share. Opportunists resume quota busting under the cover of rising oil price and booming demand. Notice that independents' capacity continues to grow over the interval 2008 to 2010 and beyond despite a steep decline in profitability of new capacity from 44 per cent (0.44) to 23 per cent (0.23) in 2010. Even so, the financial return from new upstream projects comfortably exceeds the hurdle rate and there is lots of new capacity in the pipeline from the previous decade of investment.
In the remaining years of the simulation to 2020 the relentless expansion of independents' capacity creates periodic crises within OPEC, leading the swing producer to renewed bouts of punitive production that induce volatility in oil price and demand. This instability of supply in the period between 2008 and 2020 restricts oil price to a range between $10 and $35 per barrel. Moderate, though volatile, oil price is conducive to greater consumption in a growing global economy and consequently demand rises. The resulting upstream investment climate remains mostly favourable (with the exception of a notable downturn in profitability in 2014) and so the independent producers continue to expand capacity and to grab most of the extra global demand.
Figure 8.25 shows the effect on commercial reserves of the Asian boom scenario with quota busting. Independents' undeveloped reserves (line 1) begin at 471 billion barrels in 1996 (471 000 million on the vertical scale), while risky Russian reserves (line 2), known to exist but not yet adopted by any major commercial producer, begin at 210 billion barrels (210 000 million on the vertical scale). Over time, risky Russian reserves steadily fall as confidence builds among commercial producers to exploit more and more Russian oil. These risky reserves are essentially transferred into independents' undeveloped reserves. The rate of transfer is governed by the parameter 'time to build confidence in Russia' which is set at 20 years in the scenario. By 2008 risky reserves are down to 115 billion barrels and by 2020 to 63 billion barrels, barely a quarter of the initial amount. The transfer of almost 150 billion barrels of Russian oil replenishes independents' undeveloped reserves and extends their lifetime. Nevertheless, independents' reserves decline steadily throughout the simulation because the rate of development of new commercial oil fields exceeds replenishment. Reserves fall to 329 billion barrels by 2008 and to 202 billion barrels by 2014. Notice that in the period 2014 to 2015 independents' reserves are virtually stable at 200 billion barrels despite production of 56 million barrels per day. This plateau corresponds to a brief period in which replenishment of reserves (through adoption of risky Russian oil) exactly balances new development in the wake of a dip in new upstream projects caused by low oil price. Thereafter, reserves continue to fall, declining to 75 billion barrels by 2020.
Figure 8.25. Commercial reserves and development cost in a contemporary scenario: Asian boom with quota busting, cautious upstream investment and Russian oil
Figure 8.25 also shows development cost per barrel. This trajectory merits close investigation. Development cost starts quite low at $12 per barrel in 1996. It rises steadily over the 25-year simulation to $21 per barrel in 2020. This increase is modest considering that total non-OPEC reserves have fallen 681 billion barrels in 1996 (the sum of independents' undeveloped reserves and risky Russian reserves) to 138 billion barrels in 2020. Almost 80 per cent of known commercial reserves are used over this period. Surely such extensive depletion means that remaining recoverable reserves will be much more expensive to extract. However, escalation of development cost is curtailed by two factors. First, known Russian reserves, available at modest cost, are steadily replenishing commercial reserves as independent producers build confidence in Russian oil. Second, recovery technology is advancing throughout the quarter-century scenario period, as reflected in the downward sloping technology cost curve shown on the right of Figure 8.6. If this particular technology assumption is broadly correct, then it means that recovery costs in 2020 for a given oil field are only 55 per cent of their 1996 value. So far, technology has delivered on this efficiency assumption.
8.6.4. A Contemporary High Price Scenario – How to Push Oil Price Over $60 per Barrel
In 2006, oil price was almost $70 per barrel and the pump price of petrol in the UK was very nearly £1.00 per litre. This price level is much higher than achieved in the simulated scenario above. So what scenario conditions are needed to push oil price into this realm? Apparently buoyant demand alone is not enough, although it is certainly an important scenario ingredient. The model suggests that it is difficult to invoke high oil price because OPEC oil is cheap and even non-OPEC oil can be extracted for less than $25 per barrel under quite aggressive assumptions about growth in global demand. If oil price exceeds $40 per barrel then commercial production is highly profitable and, sooner or later, adequate supply will be forthcoming. Conditions that would cause chronic undersupply are an Asian boom coinciding with restricted supply from OPEC, but is OPEC undersupply plausible in light of quota busting by opportunist producers? The answer is yes if a combination of political uncertainty, economic sanctions and military conflict in the Middle East conspire to reduce OPEC output so much that opportunist quota busting is cancelled out.
The scenario parameters in Oil World 1995 can be adjusted to capture the political uncertainty that would curb OPEC output. Two changes are required relative to the previous scenario. Opportunists' capacity bias is reduced to zero meaning that collectively opportunist producers are unable to engage in quota busting because oil exports for some are curtailed through sanctions or war. In addition, capex optimism is reduced to its minimum value of 0.5 to represent extreme caution in the investment decision of independent producers because of assumed political volatility and conflict in the Middle East. Otherwise, all the other scenario parameters remain at the same settings as in the previous scenario, including of course the assumption of an Asian boom driving global economic growth.
Figure 8.26 shows a simulation of this new scenario. Only two time charts are shown, but additional charts can be inspected by running Oil World 1995 with the parameter settings described. Oil price (line 1, bottom chart) rises steeply from $16 per barrel in 1996 to $40 per barrel in 2008, increasing still further to $63 per barrel in 2016 and $68 per barrel in 2020. There is also considerable volatility in price, with peaks of more than $60 per barrel and troughs of $30 per barrel. This volatility is caused by an interaction between the swing producer (as price regulator) and consumers. Demand growth is curtailed by high price despite the assumption of an Asian boom. Demand for oil rises from 68 million barrels per day in 1996 to 74 million barrels per day in 2020.
The simulation shows that it is possible for oil price to rise very high even when development costs remain at less than $20 per barrel (as they do in this particular scenario). High price is sustained by a combination of demand pressure and politically induced restriction in OPEC supply. Even so, the simulator cannot fully match the price peak of 2006. Most likely, this shortcoming is because the model does not fully represent oil stocks, hoarding and speculation in oil markets, short-term factors that make oil price particularly sensitive to political uncertainty in the Middle East. Nevertheless, the simulator clarifies the enduring economic and political forces driving long-term supply, demand and price in the oil industry.
Figure 8.26. Demand and oil price in a contemporary high price scenario: Asian boom with OPEC undersupply, very cautious upstream investment and Russian oil
