Is the Price of Oil Still a Good Proxy for the Price of Energy?
- Khalil Zahr
- Aug 28, 2019
- 9 min read
Highlights
The world’s primary energy consumption accelerated in 2018, confirming a rising trend that started in 2016.
The observed growth in consumption, however, is inconsistent with country-based energy models’ predictions. World GDP growth in 2018 was less than that in 2017, while crude oil prices were over thirty percent higher. In light of these facts, actual growth in energy consumption should have slowed.
While weather effects, due to climate change, may partially explain such inconsistency, it does not offer a full explanation.
A closer look at the principle energy markets and their price trends, reveals a weakening correlation between the price of crude oil, on the one hand, and the price other energy sources, on the other.
Consequently, the inconsistencies observed between actual energy consumption and model estimates, are also due to an issue with the model framework stemming from the erroneous use of the price of crude oil as proxy for the energy price.
The acceleration in fossil energy consumption over the past three years is more likely driven by an extended period of low to moderate price levels. These levels are also contributing to the slowing growth in renewable energy, to the detriment of the climate agenda.
Introduction
In a previous post in July 24, 2018 addressing the challenges posed to the climate change agenda by low energy prices, I noted that the resurgence- accelerating growth for fossil energy on the one hand and slowing growth of renewable energy on the other hand- could be an outlier and further data is needed in order to confirm this negative trend. Unfortunately, for concerned climate observers, the 2018 data not only confirms the suspected trend, but also indicate a worrying acceleration and robustness.
World primary energy consumption grew at a notable 2.9% rate in 2018, far exceeding the average annual rate of the previous seven-year period (2010-2017) of 1.7%. This trend in consumption was global, with the departure from historical trend most pronounced in the developed OECD countries, and also notable in the developing Non-OECD countries.
The observed growth in consumption, however, is inconsistent with country-based energy models predictions which rely on economic growth and crude oil prices-as a proxy for energy prices- for predicting energy consumption. World GDP growth in 2018 was less than that in 2017, while oil prices were over thirty percent higher. In light of these facts, actual growth in energy consumption in 2018 should have been much slower than in 2017.
In its 2019 Statistical Review of World Energy, British Petroleum Co. (BP) Economists stated that this development may be related to weather effects revealed in “unusually large number of hot and cold days across many of the world’s major demand centres last year”. The “weather effect” is certainly a logical explanation where the shift in the price elasticity of demand is not fully reflected in the energy model’s parameters. But this effect does not explain these models underestimating the growth in energy consumption (as also observed by BP economists) over the period 2014-2016, when oil prices experienced a sharp decline. These inconsistencies seem to indicate the presence of other factors, besides the temperature effect, have contributed to the divergence between model forecast and actual energy consumption.
By taking a closer look at the principle energy markets and the price trends, not only of crude oil, but other energy sources such as natural gas and coal, one can observe a weakening correlation between the oil price, on the one hand, and the price of those other energy sources, on the other. Consequently, the oil price is not anymore, and for various reasons, a reliable proxy for energy market developments. This weakening correlation also factors into the divergence between the energy models forecasts and actual consumption.
The implications of this evidence are important for two primary reasons: first, international energy statistics gathering, and forecasting organizations cannot rely anymore on simplified macro models to forecast country-level energy consumption for most of the major oil markets. They need, instead, to employ more detailed sectoral models that are data demanding, and time consuming. Second, with the rising concerns about climate change, it is essential to properly assess and account for its impact on energy consumption. If it turns out that climate change is driving energy consumption, specifically fossil energy, then a vicious circle could be forming, with the latter feeding back into extreme climate events and extreme temperatures.
Trends in Energy Consumption
World energy consumption in 2018 grew at high rate of 2.9 percent, well above the annual average for the past 7-year period (2010-20179) of 1.7 percent. It confirmed an accelerating trend which started in 2016, following a short period of moderating growth (Figure 1).
Notable is the growth in consumption in the developed OECD group, where it reached seven and half times the average for the period 2010-2017, (1.5 vs. 0.2 present respectively). Most of the growth in this group came from the United States, whose consumption grew at more than eight times the average. Also, the European Union (EU) group witnessed a decline of only -0.2 percent in total energy consumption compared to the above period average of -0.6 percent, signifying a marked slowdown in efficiency gains.
In the Non-OECD group of developing countries, total energy consumption grew by 3.9 percent in 2018, or by 30 percent above the historical average. Most pronounced was the growth in India and China. Indian consumption grew by almost 8 percent, well above the historical average of 5.2 percent. China’s consumption on the other hand, registered a growth of 4.3 percent in 2018 compared to the historical average of 3.5 percent (Figure 2).
The Income Factor
The observed exceptional growth in world energy consumption in 2018 is not sufficiently supported by actual economic growth which, together with energy prices, are the determinants of energy demand. World Gross Domestic Product (GDP) grew by 3.6 percent in 2018, which was below that of 2017 (3.8 percent) and in line with the average economic performance of the past eight years (Figure 3).
At the level of major energy markets, economic performance, with the exception of the US, was not consistent with observed growth in energy consumption in these markets. While the US economy grew at 2.9 percent in 2018 from 2.2 in the previous year, China, India, and the EU registered moderately slower growth in 2018 than in 2017 (6.6, 7.1, 2.1) vs. (6.8, 7.2, 2.7) percent respectively (Figure 4).
The Price Factor
If the influence of economic growth on energy consumption is at best neutral, then energy prices had to be the primary drivers of its growth. This however is not supported by actual developments in crude oil prices.
After reaching a low in 2016, crude oil prices rose by 23 percent in 2017 and 30 percent in 2018. Consequently, if the price of crude oil is a reliable proxy for the price of energy, then energy consumption should have slowed in 2018 (Figure 5).
The Price of Crude Oil vs. the Price of Energy
A detailed look at the developments in crude oil prices versus the prices of other energy sources, such as natural gas and coal, shows a growing divergence between them. This means that the price of oil is no longer a reliable proxy for the price of energy as it was in the past. Consequently, using it as such in energy modelling will cause the divergence between model estimates and actual energy consumption and explain other inconsistencies. While increased number of hot and cold days, due to climate change, explain some of these inconsistencies, a well-defined modelling frame work will not produce such estimation errors. The impact of climate change may modify consumer behavior, thus shifting the price and income elasticities of energy demand but will not change the fundamental drivers of consumption: income and price.
A closer look at energy price developments in the major world markets clearly reveals the weakening correlation between the price of crude oil and that of other primary sources of energy, such as natural gas and coal, particularly in the advanced economies.
The US Market
In the US market for instance, while the price of West Texas Intermediate (WTI) (the marker crude oil), rose from under 14 USD per million British Thermal Units (BTUs) in 2010 to over 16 USD per million BTUs in 2012. In contrast, the price of natural gas decreased from over 4 USD to about 3 USD per million BTUs over the same period. The price of coal on the other hand stayed within a narrow band around 3 USD per million BTUs. Between 2012 and 2014, crude oil prices did not appreciably change, while those of natural gas rose, and the price of coal declined. In the following period; 2014-2016, oil prices sharply declined by over 50 percent, while the prices of natural gas and coal followed suit but at much slower rate. When energy prices recovered over the period 2016-2018, crude oil prices rose by over 50 percent, while the prices of natural gas and coal rose by 27 percent and 47 percent respectively (Figure 6).
Consequently, the exceptional growth in US energy consumption in 2018 is a product of moderate prices for natural gas due to increased supply of shale gas and relatively low coal prices pressured by lower demand for coal fired power plants, primarily driven by environmental concerns (Figure 7).
The EU Market
As in the US, divergence between crude oil prices and the prices of natural gas and coal is well pronounced, particularly with the latter. Prices of coal did not follow the steep variations in crude oil prices over the past 8 years. Natural gas however has tended to follow the price of crude oil for the first half of the 2010-2018 period, while exhibiting wide divergence during the second half (Figure 8).
The Chinese and Indian Markets
The Chinese and Indian energy markets exhibit similar price dynamics and trends. Both countries import liquified natural gas (LNG), whose prices usually track the prices of fuel oil. Consequently, and unlike the cases of the US and EU, the price of natural gas exhibits high degree of correlation with the price of crude oil. This, however, is not the case for coal which exhibits weak correlation with the latter. Consequently, even in these two major markets of the developing world, the use of the price of crude oil as a proxy for the price of energy will introduce sizable estimation and forecasting errors. Moreover, the growing use of nuclear energy as an energy source in these markets, and in light of the decoupling between the price of Uranium 308 (the fuel for nuclear power plants), and that of crude oil, further weakens the latter’s role as proxy for total energy (Figures 9&10).
A Questionable Energy Model Framework
The observed inconsistency between actual energy consumption and what a macro-energy model would have predicted, motivated BP economists, the authors of the 2019 BP Statistical Review of World Energy, to look further into its possible causes. They concluded that this observed inconsistency “may be related to weather effects. In particular, there was an unusually large number of hot and cold days across many of the world’s major demand centres last year, particularly in the US, China and Russia, with the increased demand for cooling and heating services helping to explain the strong growth in energy consumption in each of these countries”.
By augmenting the model to include a measure of heating and cooling days for countries, for which data was available, BP economists were able to “greatly reduce the extent of the surprise in last year’s energy growth”. But the growth in energy demand still looked stronger than expected. Furthermore, these factors do little to explain “the surprising weakness of demand growth in the period 2014-16, which is far lower than the framework predicts”.
The “weather effect” notwithstanding, it seems that the divergence between actual energy consumption and the energy model’s estimates are due to the model framework. Specifically, the use of the price of crude oil as a proxy for the price of energy is no longer appropriate for many of the world energy markets. This is due to the growing divergence between the price of oil on the one hand, and the prices of other energy sources such as natural gas, coal, and nuclear fuels. More detailed models are needed for these markets, which incorporate the specific dynamics defining the demand for the individual energy sources.
Conclusions
The acceleration in fossil energy consumption over the past three years is more likely driven by an extended period of low to moderate price levels. These levels are also contributing to the slowing growth in renewable energy, to the detriment of the climate agenda. Carbon dioxide emissions, in line with the growth in fossil energy use, have accelerated over the past three years, after a short period of moderating growth (Figure 11).
As I have emphasized in a number of posts on this blog, slowing and reducing greenhouse gases (GHG) emissions will require putting a price on them that reflects the economic and social costs of global warming. This goal, however, cannot be achieved without cooperation at the global level. Unfortunately, the necessary cooperation is still well below what is required to limit global warming below 2 degrees centigrade, as called for by the Paris Accord on climate change.
END
Comments