Michael C. Lynch
Energy Policy Research Foundation
Washington, DC
It seems likely that, absent major climate-change policies that reduce oil demand significantly, the crude oil market should be less volatile than in the past two decades, but hardly stable. A moderate amount of surplus capacity in Organization of Petroleum Exporting Countries (OPEC), a less stressed downstream sector, and ample commercial inventories should translate into moderate price cycles. However, should expansion plans in OPEC members go forward, and especially if Venezuela is able to revive its upstream sector in the next few years, there will be a growing surplus of crude production capacity, and potentially new price wars. A political supply disruption could also always send prices much higher.
Should the energy transition gather momentum, lower demand could mean weaker prices and more frequent price wars, but if the remaining production becomes concentrated in the Middle East and Russia, market stabilization efforts would prove much easier. But having supply increasingly concentrated in national oil companies could mean less timely decision-making, prompting more frequent mismatches between supply and demand and the price spikes that come along with it.
Overall, as with other commodities, oil prices are likely to remain volatile for some time, and the longer trend will depend very much on political decisions made by both producing and consuming governments. Oil companies would do well to remain flexible in their planning and operations and avoid over-reacting to short-term price movements.
Background
In 2001, my article “Oil prices enter a new era of increased volatility” appeared in Oil & Gas Journal, arguing that a variety of factors would make oil prices more volatile than in the previous decade, and this would translate into higher average prices: $30/bbl, adjusted for inflation. Needless to say, that was a significant undershoot, with average prices over the past 20 years at roughly $75/bbl. But the view about long-term prices was not inconsistent with the then-consensus (Fig. 1). This certainly supports the argument that forecasters prefer to embrace the consensus, or perhaps fear being the outlier.1
But long-term forecasts do not address the issue of short-term volatility, which is not surprising as many of the factors that affect volatility are both difficult to predict—weather, gross domestic product (GDP) growth, OPEC policies—and less relevant over the long-term (recessions, war, sanctions).2 Monthly oil demand fluctuates significantly more than annual demand, but these fluctuations are irrelevant to prices over the course of decades.
Still, short-term price movements are important for a variety of reasons. Volatility creates uncertainty about revenue for the industry, which can make planning difficult and raise borrowing and investment costs, especially for smaller players. And not infrequently, sudden price moves are interpreted as representing changes in long-term trends, as when the 1979 Iranian Revolution convinced nearly everyone that prices would continue soaring, or the way the 1998 oil-price collapse saw some companies selling out to competitors. This has often led to poor strategic decisions and significant financial losses.
The expectation of higher price volatility was based on several factors: tighter capacity, including OPEC crude production, but also in the refinery sector and tanker market; a reduction in private oil stocks as the industry sought to be more efficient; product regulations that increased market barriers; and a decrease in the timeliness and reliability of data. Many blamed poor data for price volatility and the Joint Oil Data Initiative was created in an effort to improve that situation.3
But the past two decades have seen enormous swings in prices, almost entirely due to external events: the second Gulf War, the decimation of Petroleos de Venezuela’s workforce, the Arab Spring, and sanctions on Iranian oil exports all pushed prices up, while the 2008 financial crisis and the 2020 pandemic caused them to crater. (Fig. 2) Low, even minimal, spare capacity in OPEC made it difficult for the group to prevent the price spikes, though they proved quite capable of reversing the two major price collapses in 2008 and 2020.
Obviously, the geopolitical problems that caused much of the price volatility of the past two decades have not disappeared, but there is every possibility that they will be largely mitigated in the coming decade. The restoration of oil exports from Iran and Venezuela, while uncertain, seems likely over that period, which should mean much less market tightness, assuming other major producers do not restrict production. The recent move by the UAE to have OPEC and its allies (OPEC+) recognize its increase in production capacity and the Saudi economic diversification program suggest that producers are now focused on increasing oil revenue, which will almost completely be due to higher sales volumes, not prices.
The rise of shale oil production, presently dominated by the US, is also a constraint against raising prices too high. The oil shale resource is enormous, but costs more than much conventional oil production especially in the Middle East. This means that shale oil supply is more responsive to prices, particularly when they are close to break-even costs. Fig. 3 shows the change in shale oil production compared with prices, and generally below $50/bbl production declines.
Market fundamentals that normally increase or reduce price volatility have changed due to the pandemic, but their status in the post-pandemic world remains uncertain. Arguments that oil demand has or will soon peak, which received more attention due to pandemic lockdowns, seem less valid now that the pandemic impact is easing and most oil demand has returned to pre-pandemic levels, with jet fuel being the exception.
Capacity utilization
The availability of spare capacity in various segments of the industry has always been an important factor in determining price volatility, although the importance varies over time and by sector. The price spike in the 1970s and the subsequent demand slowdown created major surplus capacity throughout the industry, such that price volatility was relatively low during the 1990s. But that changed as the market moved into equilibrium, especially after the 1998 oil price collapse. The pandemic has meant a major increase in surplus capacity, but much of this is certain to be temporary.
The upstream outlook is somewhat muddled, partly because there is often disagreement about capacity and especially the amount not being produced. Iranian production has been depressed by sanctions for several years, and it is not clear if the pre-sanctions capacity estimate remains valid. In theory, however, maintenance and some investment would allow full capacity to be restored, after a delay. Also, some have questioned whether Saudi ‘strategic surplus capacity,’ which is said to be about 2 million b/d is actually available. Others note the spare capacity is for heavier crudes, making it less useful as regards market tightness.
The 2008 recession meant a major increase in OPEC surplus capacity which was slowly worked off (Fig. 4), then the rise in shale oil production pushed down demand for OPEC oil. Subsequently, a loss of capacity in Libya, Nigeria, and especially Venezuela offset that. In the months before the pandemic began, OPEC surplus capacity was quite small, although there was a small amount held by non-OPEC producers coordinating with OPEC, primarily in Russia.
The end of the pandemic should thus see a return to the minimal levels of surplus capacity that prevailed 2015-19, translating into more price volatility. Longer term, the return of Iran or Venezuela, plus expansion of capacity in Iraq, Kuwait, and the UAE, could mean periods of weak prices, especially as none of those additions are price driven. Lumpiness in supply from those three countries might thus cause prices to be more volatile, primarily downward.
As discussed later in this article, more aggressive climate change policies could also mean restrictions on investments by non-state oil companies, especially in the OECD, as well as reduced demand. The former would mean tighter supply, all else being equal, and thus volatile and higher prices, while the latter would increase spare capacity and cause prices to weaken, with OPEC+ waging a constant battle against price dips.
The midstream and downstream sectors have traditionally not been drivers of crude price volatility, partly because shipping and refinery costs are a relatively small part of the delivered price of crude. Even so, periods like 1996 and 2000 saw tight product markets helping pull up crude oil prices, as refiners bid for the higher-quality supply.
Still, Fig. 5 shows that capacity utilization has declined from its 1995-2005 highs, despite having recovered from the impact of the 2008 recession. Non-OECD capacity utilization tends to lag as some countries find technical and financial problems keep refineries operating well below optimal levels in some countries. But even OECD utilization has been lower than in the heyday of the 1990s, when tight refinery markets meant higher prices for light products, which in turn increased differentials between light and heavy crudes.
Flat or possibly declining OECD demand will mean capacity utilization remains low in those countries, as refinery closures lag demand trends, whereas non-OECD demand can be expected to outpace refinery investments, especially given financial constraints due to pandemic-inspired government debt levels. Some governments, such as Mexico’s, do push for refining investment regardless of the economics, so it is possible that the sector will continue to have low capacity-utilization levels even in the non-OECD countries, which should not affect crude oil price volatility, but could reduce differentials between low- and high-grade crudes.
Inventories
Oil inventories have long allowed the industry to cope with minor fluctuations in demand, seasonal or otherwise. As with any commodity, prices tend to be inversely related to inventory levels but, more importantly, higher inventories provide a cushion to deal with unusual events, although they are rarely enough to cope with major supply disruptions. As mentioned, the industry adopted a strategy of just-in-time inventories in the mid-1990s, which might have contributed to higher prices in the 2000s. But supply disruptions in Iraq and Venezuela probably played a far larger role.
Since the 2008 financial crisis, commercial inventories have tended to be significantly higher, even ignoring the pandemic impact (Fig. 6). Arguably, this reflected the greater volatility in supply, given the Arab Spring and on-again, off-again sanctions against Iran which created a perception of less secure supplies. Lower interest rates might have meant lower borrowing costs and made inventories cheaper to hold, but the high oil price during most of this period would have offset that.
In the future, higher inventory levels should theoretically contribute to lower price volatility, at least on the upward side, but only in response to short-term market imbalances. During periods of surplus, higher inventory levels enable buyers to hold off purchases, exacerbating price weakness. In the longer run, should the overall political situation in oil producing nations improve, then inventories are likely to decline again, and this could mean greater volatility on the high side.
Product demand
The oil crises in the 1970s might have been worse if it were not for the fact that cheap oil before 1974 resulted in heavy consumption of residual fuel oil, especially for power generation. Residual fuel oil is the most price sensitive of the oil products, and OECD consumption of it dropped relatively sharply, by 1.4 mb/d from 1973 to 1975, whereas gasoline consumption rose slightly. In subsequent years, natural gas, coal. and nuclear power backed out most resid consumption, which now represent a mere 7% of the global barrel (Fig. 7), while light and middle distillates comprise 65% versus 56% in 1973. Other products’ share, primarily industrial oils and petrochemical feedstocks, rose from 15.5% in 1973 to 27% last year.
Although price responsiveness of demand fluctuates according to infrastructure, technology, and fuel and product availability, it is generally believed that gasoline is the least responsive and residual fuel oil the most, perhaps three times as responsive as gasoline. This means that a higher demand share for light products requires much higher prices to reduce demand when markets are tight or a supply crisis hits. Middle distillate and jet fuel fall somewhere in between, but also respond more strongly to economic growth rates than gasoline does.4
Fig. 8 shows projections for oil use in the transport sector under different International Energy Agency (IEA) scenarios. The breakdown of usage for road, shipping, and aviation does not include types of fuels for each, but road travel tends to dominate energy usage. Interestingly, the IEA anticipates little change except in the two more extreme scenarios, in which electric vehicles (EVs) come to dominate vehicle stock and oil usage for transport drops.
It is unlikely this will mean greater demand price-responsiveness, as most of the remaining oil use would be in industrial applications, such as petrochemical feedstocks. Industry tends to be more price responsive in general, and other fuels such as natural gas can substitute for at least some processes. But specialty chemicals such as lubricants, waxes, and asphalts are probably more difficult to replace.
OPEC+
It is generally argued that OPEC has served to stabilize prices,5 with the most significant price spikes due to political events beyond its control. But at least two of the price collapses (1986 and 1998) were due to conflicts within the group. The organization has definitely shown much greater cohesion since the early part of the century, however, with adherence to quotas generally high. The most serious recent challenge came in early 2020, before the severity of the pandemic was obvious, when Russia, which had been cooperating with the group, initially resisted continuing to do so.
The pandemic obviously overwhelmed the industry, with an unprecedented collapse in demand reaching 16 million b/d by second-quarter 2020. Prices dropped alarmingly, below $20/bbl in April 2020, and Russia and a number of smaller producers agreed with OPEC to reduce production by 10 million b/d, bringing the market back into balance. Since them, compliance with quotas has been high, often over 100%, but partly because some producers like Angola and Nigeria have had trouble meeting their production targets.
The future abilities of the organization and its allies remains uncertain. Iraq, Kuwait, and the UAE all plan to expand production, and the Russians are likely to increase their capacity, albeit more gradually. In the long run, Iran and Venezuela are likely to return to the market, adding millions of barrels per day of supply, although apparently not anytime soon. Still, the potential for the group to face future battles over quotas and sustainable pricing should not be underestimated.
Energy transition, market evolution
Once again, the petroleum industry is facing arguments that its days are numbered or that at least oil demand will soon peak because of climate change policies and technological advances, mostly in electric vehicle batteries. Peak oil demand as a concept is much more valid than peak oil supply (no mineral has ever undergone a peak in supply due to resource constraints), but as always, caution must be exercised given the many times it has been predicted in the past. The Rocky Mountain Institute’s report “Winning the Oil Endgame,” opens with numerous quotes from industry executives suggesting that hydrogen, fuel cells, or other technologies would bring an end to the oil era,6 with a skeptical counterview published by Lynch and Sandrea.7
Assuming the transition proceeds, there are a number of implications for price levels and volatility. Fig. 9 shows the price forecasts for the IEA’s varying scenarios from the 2021 World Energy Outlook. The IEA sees oil prices as being positively correlated with demand—lower demand equals lower prices—because demand trends are not driven by prices but climate change policies. Pushing demand down means oil prices would be lower. IEA’s price in the Stated Policies scenario appears to be much too high. Historically, its price expectations have erred on the high side.
This is important because if prices decline significantly in the future, more expensive sources of oil, such as US shale, will see less investment. Lower non-OPEC production more generally would see the market share of Russia and especially the Middle East increase. These dynamics could be exacerbated by political pressure on private oil companies to reduce investment in oil production.
Interestingly, IEA does not project a significant change in OPEC’s market share in its most recent scenarios, except for NetZero 2050, in which it anticipates that the market share for OPEC+ will increase by 2050 to 61% from 47%. While it is possible that political steps, such as oil import tariffs in the US, could prop up non-OPEC production, there is a significant chance that OPEC, with or without the members in OPEC+, will see growing market share over the long-term, perhaps again approaching 50% as the energy transition proceeds. This growth in OPEC market share would be particularly pointed in the event of either a dead stop to or phasing down of non-OPEC investment (Fig. 10).
If, as IEA projects, global oil demand drops to 25 million b/d by 2050 in its NetZero2050 scenario, it is possible that much of that oil will come from Saudi Arabia and Russia. Even in the Sustainable Development Scenario, in which oil supply drops to 47 million b/d by 2050, half of that could be from Russia and Saudi Arabia.
In theory, this could result in much less volatile oil prices, not unlike 1975-85 (Fig. 3) when OPEC countries were price-makers, not price-takers. The type of market management seen during and before the pandemic, which was extremely effective, would be far easier with a duopoly or a group of exporters having the bulk of the market share. A quick phone call could see prices set for the coming quarter, and the rest of the industry would simply follow suit.
Which doesn’t mean there wouldn’t be periods of extreme volatility. Setting prices too high could result in a market collapse, as happened in 1986 or 2014. At this point, it appears more likely that shale production costs will drop more slowly in the future than over the past decade, meaning that only significantly higher prices would result in another production boom such as caused the 2014 price collapse.
But there is also the problem of concentration of supply in potentially volatile regions. To date, none of the drone and missile attacks on Saudi oil fields have done serious damage, but there is certainly a serious threat to future operations in a region that has seen four major wars in four decades, as well as a number of smaller conflicts. Russia has its own history of revolution and coups, along with no small amount of regulatory risk. Beyond the oil market’s own behavior, the pandemic has shown the risks related to relying on a few large suppliers.
Another aspect that could affect price volatility involves increasing share of the market supplied by national oil companies, and the attendant expanded role of governments in determining upstream investment levels. Some state oil companies, like Saudi Aramco, appear to plan investments according to expected demand. Others, like Pemex, suffer from constant fluctuations in their budgets based on government liquidity (and politics). Kuwait Petroleum Co. has seen proposed investments blocked because of political disputes in the parliament, and the Iraqi government has had its upstream plans repeatedly delayed by internal discord. The continuation of such events could mean greater price swings in the future as supply increments are increasingly delayed by domestic political disagreements in exporting countries, as well as by greater volatility in upstream expenditures as they become increasingly unrelated to either price levels or expected demand.
Renewables, BEVs
If optimistic projections for the growth in renewable power and battery-electric vehicles (BEVs) prove true, then the impact on oil price behavior could be significant. Although solar and wind provide electric power, not transportation fuels, their variability could have an impact on oil and gas prices as regards back-up generating capacity. In most cases this capacity is provided by gas turbines. But because weather can cause a surge in demand for natural gas, and long-distance transportation of it is expensive and requires large-scale investment, there could be a growing need for distillate fuel oil to power dual-fired turbines.
An illustrative example occurred in 2004, when China’s rail system proved unable to deliver enough coal to power plants to meet booming electricity demand and Chinese oil demand surged by 1 million b/d (about 15%), although not all was for power generation. In countries like India, where electric power is unreliable, many consumers own diesel generators for backup. Globally, the diesel generator market in 2017 was 70 GWe. It is projected to reach 100 GWe by 2022.8 Although undoubtedly much of this demand is for industrial sites, such as construction, the amount is about 1/3 of 2019 capacity additions in renewables.9
Similarly, natural gas turbines generate much of the world’s electricity. About 21 tcf of gas generates about 23% of global electric power.10 The oil equivalent is 9 million b/d. In theory, a significant fraction of this capacity is dual-fired. About 25% of gas capacity in the US can use petroleum,11 but it’s not clear how representative this is of the rest of the world.
Still, there is the potential for occasional oil demand spikes from power generators in the future, as seasonal and even short-term fluctuations create demand for backup power that can’t be met by natural gas. Should this occur when oil markets are already tight, prices could soar, not perhaps to the extent that natural gas prices have in Europe and Asia recently, but definitely more than is currently normal.
On the other hand, an increased stock of BEVs could have a moderating effect on oil-price volatility, as drivers relying on BEVs for short-distance commuting while using an internal-combustion engine (ICE) vehicle for longer trips, as many apparently do now,12 vary usage at any given time depending on gasoline prices. This can be seen to a degree in historical data. Fig. 11 shows the change in vehicle efficiency in new car sales and average efficiency in the entire fleet as used. The data is not very precise, but in 2008, for example, efficiency for the existing fleet grew by about three times that of the new car sales, implying people were driving their more efficient vehicles more often than before.
In California, BEVs seem to be used less than half as much as the average vehicle, suggesting they are often treated as second cars or used for daily commutes rather than longer trips. This means that, in times of high gasoline prices, consumers could increase reliance on their BEVs, reducing gasoline consumption. This would have a moderating impact on oil prices, but it will be many years before the fleet of BEVs is large enough for the effect to be noticeable.
References
- Lynch, M.C., “Bias and Theoretical Error in Long-Term Oil Market Forecasting,” Advances in the Economics of Energy and Natural Resources (Moroney, J.R., editor), JAI Press, Stamford, Conn., 1994.
- Lynch, M.C., “The Peak Oil Scare and the Coming Oil Flood,” Praeger, Westport, Conn., 2016.
- Lynch, M.C., “Drivers of Oil Price Volatility,” Journal of Energy and Development, Vol. 28, No. 1, Autumn 2002, pp. 107-141.
- Dahl, C. and Sterner, T., “Analyzing gasoline demand elasticities: A Survey,” Energy Economics, Vol. 13, No. 3, 1991, pp. 203-210.
- Al-Khateeb , L., “Long Live the Price Maker,” Brookings, July 5, 2015.
- “Winning the Oil Endgame,” Rocky Mountain Institute, Snowmass, Colo., 2005.
- Lynch, M.C. and Sandrea, I., “The Pandemic and the End of Oil?”, EPRINC Working Paper, February 2021.
- GlobalData Energy, “Global diesel generators market to reach value of $115.5bn in coming years,” power-technology.com, Feb. 14, 2018.
- International Energy Agency (IEA), “Renewable Energy Market Update: Outlook for 2021 and 2022,” 2021, p. 6.
- IEA, “World Energy Outlook 2021,” Table A-3, October 2021.
- US Energy Information Administration, “Electric Power Annual,” Table 4-11, Oct. 29, 2021.
- Burlig, F., Bushnell, J.B., Rapson, D.S., and Wolfram, C., “Low Energy: Estimating Electric Vehicle Electricity Use,” National Bureau of Economic Research Working Paper 28451, February 2021.
The author
Michael C. Lynch is president of Strategic Energy and Economic Research, a Massachusetts-based consultancy and a Distinguished Fellow with Washington-based Energy Policy Research Foundation. He earned an S.B.-S.M. from the Massachusetts Institute of Technology (MIT) political science department in 1979, and afterwards held research positions at MIT. Subsequently, he was chief energy economist at DRI-WEFA, and served as president of the US Association for Energy Economics. He is a senior contributor at forbes.com, and his publications have appeared in seven languages. Praeger published his book ‘The Peak Oil Scare and the Coming Oil Flood’ in July 2016.