Paradoxical? An Analysis Of Seasonal Trends In The Natural Gas Sector
Sunday, March 3, 2024
Welcome to the first in a series of commentaries I plan to write about topics in the oil and natural gas sector. These will typically be published on Saturday or Sunday and will be publicly available. I hope you find them interesting and educational.
As much as I talk about the weather on this site, you might think that natural gas prices will perform best when demand is high—i.e. during the winter. Interestingly, this has not historically been the case and can often result in a bear trap for inexperienced investors. This article will detail seasonality trends for natural gas and speculate on explanations for them.
For the period of evaluation, I used 2009-present as 2009 kicked off the modern era of natural gas fundamentals, characterized by a persistent over-supply as fracking technology sent production surging while demand growth lagged.
Getting straight into it, the 15-year average year-to-date price change is shown in the Figure to the right.
There are a couple of takeaways from this data. First, natural gas prices have not seen significant growth with the commodity averaging a annual -2.0% loss. Prices started 2009 at $5.97/MMBTU as the 2008 Commodities Bubble was still in the process of bursting and finished 2023 at $2.51/MMBTU. Ouch. Natural gas is cyclical, not a buy-and-hold growth commodity, such as gold.
Second, and more importantly, one can identify four broad trends over the course of the year. I will break these up into 4 phases.
1) Late Cooling Season & Fall Shoulder Season: Prices begin a choppy uptrend from early August peaking for the year in mid-to-late November as average year-to-date gains top +12%.
2) Heating Season: Prices then begin a rapid pullback, giving back all of their gains before the end of the year. This sell-off extends into the next year, ultimately nearing -25% by the time prices bottom in March.
3) Spring Shoulder Season: After establishing a baseline, prices then trend steadily higher until early June, rallying on average 12%-15%.
4). Early Cooling Season: Prices then experience a quick pullback before stabilizing from early June into early August ahead of the next late-year run-up.
So, what is going on here?
Starting with phase 1: The August-November choppy rally—which encompasses the late cooling season and entire Fall Shoulder Season—occurs as investors look towards the coming winter and speculate on the impact of an early-season arctic outbreak on the adequacy of storage for the rest of the heating season. While actualy demand during this time is near annual lows, the uncertainty about the coming demand maximum tends to drive prices higher. Between August 15 and November 30, prices climb an average of +7.8%.
Prices then have a tendency to sell-off in Phase 2 during the winter. While this is typically the peak of heating demand when natural gas sees the largest storage withdrawals, inventories are invariably sufficient to meet demand and this becomes clear early on, generally within the first month of the season, though some exceptions, such as the Winter of 2013-14, do exist. This sell-off extends throughout the season with prices bottoming in March. That is not to say that weather is not important here. A big time arctic blast—such as we saw this past January—can send prices spiking. However, it takes sustainable cold and tight inventories to keep prices up. Even consistent, seasonably cool temperatures is still probably not enough to protect the premium built into the commodity during the Fall. Between November 30 and March 15, prices drop an average of -16.9%. Ouch.
Phase 3 consists of the late Heating and Spring Shoulder Seasons when prices have become oversold after the winter drop and investors turn their attention to the upcoming Summer Cooling Season. Even if inventories are bloated, there is nearly 7 months before the next Heating Season—plenty of time to fix things—which leads to dip buyers stepping in. Between March 15 and June 10, prices rally an average of +14.1%. The best-performing year for this period was 2022 when prices soared by +92.2%--and they might have kept going had not a fire at the Freeport LNG export facility promptly knocked over 2 BCF/day of demand offline for the next 8 months.
Finally, by the middle of June, if the heat hasn’t shown up and storage levels are still robust, prices begin to stagnate and even pullback as investor optimism wanes, as shown in Phase 4. Between June 10 and August 15, prices on average decline a lackluster -2.2%.
The cycle then begins anew with prices ramping up through the Fall ahead of the next heating season.
On a month-by month basis, September has the best performance with an average +6.5% gain, followed closely by April at +5.0%. On the other hand, December (-6.8%) and February (-5.9%) are the underperformers. This is shown in Figure 2 above.
These trends bode well for natural gas prices moving forwards with Phase 3 set to begin in the next few weeks. The favorable seasonality coupled with lower production should argue for higher prices, even with storage surpluses at 8-year highs.
How reliable is this trend? For the Phase 3 that we are about to enter, 11 out of the last 15 years, or 73% of the time, registered a gain. In these 11 years, gas prices saw an average return of +20.3%. In the 4 down years, the average loss was only -6.7%. Thus, not only is the probability of a gain in the bullish trader’s favor, but so is the magnitude of a gain versus a loss.
One final note, these trends are based on front-month natural gas prices. Unfortunately, traders of natural gas ETFs are unlikely to actually realize the same magnitude of gains as discussed above. Why? The natural gas futures strip spends the majority of its time in a state of contango in which each subsequent month trades at a higher price than the last. ETFs like 1x UNG and 2x BOIL must sell the front-month every month or every other month and buy the next contract at a higher price. This results in a chronic underperformance, which is well-documented. As a result, the same chart as Figure 1 plotted for UNG becomes Figure 4 below:
The gains are blunted and the losses are magnified. For the Spring rally of Phase 3, the +14.1% futures gain becomes a +4.5% gain. The-16.9% winter sell-off of Phase 2 became a -26.6% drop.
Natural gas ETFs, in most situations, are not effective as a buy-and-hold tool over the long-term.
Despite this underperformance, I still feel that an understanding of seasonality is important for natural gas traders. In fact, it forms one of the components of my Fair Price Model, in addition to storage, supply/demand balances, and global pricing. There are, of course, exceptions—Winter 2013-14 saw a counter-seasonality rally into late February while Spring 2020 saw a sell-off due to COVID-19 and record warmth—but buying into the heating season or shorting into the late Summer is typically a poor idea. These trends should not be traded in a vacuum but coupled with a firm understanding of fundamentals and meteorologist to increase the profitability of your trades.
Good luck out there.
Celsius Energy
Disclaimer: Natural Gas & Oil Storage Projections, Intraday Natural Gas Stats, Renewable Energy Stats, Morning Reports, and fundamental pricing models are released by Celsius Energy as experimental products. While they are intended to provide accurate, up-to-date data, they should not be used alone in making investment decisions, or decisions of any kind. Celsius Energy does not make an express or implied warranty of any kind regarding the data information including, without limitation, any warranty of merchantability or fitness for a particular purpose or use. See full Privacy Policy HERE.
