Column: Oil and Gas Mini-Slump All Part of a Cycle
Oil and gas prices are in another mini-slump, nearly three years after they were hit by the first wave of the COVID-19 pandemic in North America and Western Europe.
But the latest downturn is part of a cycle in manufacturing activity and energy prices that has repeated with an average duration of three to four years since the early 1990s.
After adjusting for core consumer prices, the price of Brent crude has fallen by 34% from the peak in May 2022 and U.S. Henry Hub natural gas is down by 73% from the peak in August 2022.
In real terms, oil prices are in the 67th percentile for all months since 1990, down from the 86th percentile in May, while U.S. gas prices have slumped to only the 3rd percentile, down from the 86th percentile in August.
Prices have fallen in response to a combination of factors, including the impact of sanctions on Russian exports, a milder-than-normal winter, the explosion at Freeport LNG, and a slowdown in manufacturing and freight transport.
In the late nineteenth and early twentieth centuries, researchers identified several cycles in business activity, prices and interest rates:
- Kitchin cycles lasting 3 to 4 four years, attributed to the accumulation and liquidation of excess inventories (“Cycles and trends in economic factors”, Kitchin, 1923).
- Juglar cycles lasting 7 to 11 years, attributed to investment in longer-lived fixed assets such as machinery (“Commercial crises and their return in France, United Kingdom and United States”, Juglar, 1862).
- Kuznets cycles with a duration of 15-25 years, attributed to construction, demographics and migration (“Secular movements in production and prices”, Kuznets, 1930).
- Kondratieff waves lasting 45-60 years, attributed to the diffusion of major new technologies such as the internal combustion engine and electricity (“Long waves in economic life”, Kondratieff, 1926).
Researchers hypothesised cycles of differing durations could be nested, for example each 7-11 Juglar cycle could be decomposed into two or three separate 3-4 year Kitchin cycles.
In practice, both the magnitude and duration of short and long-term cycles proved too variable to be much use in forecasting (“Approaches to the business cycle, Zarnowitz, 1988).
And the expansion of the service sector, which is more stable than manufacturing, contributed to the marginalisation of business cycle research.
As a result, research on business cycles moved in other directions, and policymakers increasingly aimed to eliminate cyclical instability altogether.
Notwithstanding problems identifying, classifying and explaining cycles in output, employment, prices and interest rates, the time series for manufacturing activity continues to show a strong cyclical component.
Since 1995, there have been eight cycles in U.S. manufacturing activity, based on the Institute for Supply Management’s purchasing managers’ index, averaged over 12 months to smooth some of the short-term volatility.
Troughs around October 1995, September 1998, May 2001, February 2007, January 2009, November 2012, December 2015, November 2019 and tentatively January 2023 have occurred on average every 41 months, with a range from 23 to 69 months.
Since 1995, there have also been eight cycles in oil and gas prices, based on the change in real prices compared with the prior year and averaged over 12 months.
Oil and gas cycles have been closely correlated with each other and with U.S. manufacturing activity.
On average, troughs in oil prices occur within ±3 months of a turning point in U.S. manufacturing activity, while troughs in gas prices occur within ±4 months.
U.S. manufacturing activity appears to be forming a trough at present, with the ISM index falling below the 50-point threshold dividing expansion from a contraction every month between November 2022 and January 2023.
The current manufacturing cycle is 34 months or 39 months long (depending on whether the trough is dated to April 2020 or November 2019), approaching the average duration of 41 months for cycles since 1995.
Some softness in manufacturing activity as well as oil and gas prices should therefore be expected at this point.
Not every cyclical slowdown in manufacturing turns into a full-blown recession; there have been eight manufacturing cycles since 1995 but only three were declared recessions by the U.S. National Bureau of Economic Research.
The others were mid-cycle slowdowns, often termed a “soft patch” by policymakers, followed by a re-acceleration of activity and an extension of the business cycle.
There is no way to determine in advance whether the current manufacturing slowdown will turn out to be a mid-cycle one or a cycle-ending recession. But the type matters enormously.
If the current slowdown proves to be a mid-cycle soft patch, gas and especially oil prices are likely to rise strongly later in 2023.
Global inventories of petroleum, especially the most cyclically sensitive components such as distillates, are still below the long-term average.
In the event the economy re-accelerates, inventories will deplete quickly, and there is little spare capacity to rebuild them in the short term.
Gas inventories are currently more comfortable after a mild winter in 2022/23 but could also deplete quickly if the economy accelerates and winter 2023/24 reverts to more average weather.
By contrast, if the current slowdown turns into a cycle-ending recession, both gas and oil stocks will accumulate and prices will come under more pressure in the near term.
The resulting accumulation of inventories and spare production capacity would create some cyclical slack and defer the onset of the next upswing in prices until 2024.
In reality, the trough identified in November 2019 occurred around April 2022 during the first wave of the pandemic, but the rebound was so strong it has moved the calculated trough in the 12-month average earlier.
Reuters by John Kemp, March 7, 2023