ProShares Ultra Bloomberg Natural Gas: Not A Buying Opportunity (NYSEARCA:BOIL)

The ProShares Ultra Bloomberg Natural Gas (NYSEARCA:BOIL) is an exchange-traded fund (ETF) that’s designed to produce twice the daily performance of the Bloomberg Natural Gas Subindex. In turn, the Bloomberg Natural Gas Subindex tracks the price of natural gas futures traded on the New York Mercantile Exchange (NYMEX).

I covered BOIL on Seeking Alpha back on July 28, 2023,” rating the ETF a “Strong Sell.”

Last summer, I wrote that while I’m bullish on natural gas over the long haul, the outlook for the commodity appeared mixed at best over the short-to-intermediate term. Investors tend to overestimate the bullish demand pull from hot summer weather and associated cooling demand; winter heating demand is a far more important driver for natural gas.

I also explained why I believe BOIL is a flawed ETF for anyone looking to speculate on trends in natural gas prices over holding periods of more than 5 months.

That’s primarily due to two issues – the cost of rolling futures contract exposure over time and compounding risk resulting from BOIL’s target of tracking twice the daily change in the Bloomberg Natural Gas Subindex rather than returns over longer holding periods.

Here’s a look at returns from BOIL since July 28, 2023:

As you can see, BOIL initially rallied to an August 9th closing peak of $77.23; however, it’s now down over 71% from the close on July 28th, the day the article was published.

After a plunge of that magnitude, it’s only natural to wonder if gas is close to a bottom, or if BOIL might be worth buying at current depressed levels ahead of a potential recovery in gas prices into the coming summer cooling season.

In my view, the answer is no — as I’ll outline in this article, there are few durable upside catalysts for natural gas through early June. Meanwhile, as always, the ETF’s inherent shortcomings for tracking longer-term price movements suggest room for additional downside in BOIL into the summer months.

Let’s start with this:

Fooled by Seasonality

At the time of this writing, front-month US natural gas futures (the March 2024 contract) sell for around $1.85/MMBtu and the entire natural gas futures curve has shifted sharply lower since last summer:

Natural gas futures are priced for delivery every month of the year for years into the future. The three lines on my chart above show futures prices for delivery in every month of the calendar year through July 2025 priced on three different dates.

The blue line shows futures pricing on March 29, 2023, which represents the absolute low for front-month gas futures prices last year per Bloomberg – this line shows the price of gas for delivery from April 2023 through July 2025.

The orange line represents the futures curve priced on July 28, 2023, the date I penned my bearish piece on BOIL here on Seeking Alpha last summer – this line shows pricing for gas delivery in the months of September 2023 through July 2025 inclusive.

Finally, the grey line represents the current NYMEX curve – gas priced for delivery from March 2024 through to July 2025.

Two points to note.

First, back in late March, the front-month gas contract was for delivery in April 2023; March 29, 2023, was the last day of trading for April 2023 gas futures. On that date, gas sold for $1.99/MMBtu.

On July 28, 2023, the front-month contract was September 2023, not August 2023, because the last day of trading for August 2023 gas futures was July 27, 2023. On July 28th, September 2023 front-month gas futures settled at $2.64/MMBtu.

The casual observer might conclude natural gas futures prices rallied from $1.99 to $2.64 – about 32.7% — between late March and late July. However, that “rally” is a mirage, largely the consequence of seasonality in natural gas markets.

Specifically, April 2023 is a month of weak demand for natural gas in the US since it’s after the winter heating season and before the summer cooling season. Meanwhile, the September futures are priced for delivery at a time of higher demand, in the latter stages of the summer cooling season.

Take a closer look at those two curves in my chart above – the blue and orange curves representing gas futures prices on March 29th and July 28th, 2023, respectively – and you’ll see they’re almost identical.

In fact, while the front-month price of gas appears to have rallied between late March and late July that’s solely the consequence of the shift in delivery months. Indeed, per Bloomberg, the price of September 2023 natural gas futures on March 29th was $2.75/MMBtu, slightly higher than the $2.64/MMBtu price for the very same contract on July 28th.

This is not simply an esoteric point for investors interested in ETFs tracking natural gas prices including BOIL and the United States Natural Gas Fund, LP ETF (UNG). That’s because UNG and BOIL both track natural gas prices by buying NYMEX futures contracts for gas delivery and then rolling from one calendar month’s contract to another according to a pre-set roll schedule.

So, if you bought an ETF like BOIL or UNG on March 28, 2023, you were not buying natural gas at the front-month price of $1.99 and you would not have benefited from the phantom 32.7% “rally” in front-month gas prices from March 29th through July 28th I just outlined.

Specifically, according to Bloomberg, an investor who bought BOIL at the closing price on March 29th (a split-adjusted $71.60) would have been nursing a loss of almost 13.2% by the close of trading on July 28th. For buyers of UNG, the gain over the same holding period was just over 2.75%, a far cry from 32.7%.

Look at the grey curve and you’ll see similar implications for buyers of BOIL and UNG today.

Yes, the price of gas for delivery in the summer of this year is higher than the front-month March 2024 futures – July 2024 futures are currently priced at $2.36/MMBtu compared to $1.85/MMBtu for March 2024 futures as I noted earlier. However, just as was the case in 2023, investors buying BOIL today are likely to see little or no benefit from a normal seasonal rise in gas prices that’s already priced into the curve.

The second point of importance regarding the futures curves in my chart above is that the current (grey) curve is significantly lower than the futures curve either on March 29th or July 28th of last year.

So, while the front-month price of gas today (March 2024 futures) is only a bit lower than the front month price of gas in late March 2023 (April 2023 futures), the current curve is meaningfully lower through the middle of next year. For example, in late March of last year, the August 2024 gas futures sold for around $3.53/MMBtu, more than a dollar above the current quote of $2.43/MMBtu for the same contract.

Simply put, the near-term supply and demand balance for natural gas – into the summer of 2024 – is materially worse than was priced into the curve in the spring of 2023, a little under a year ago.

And that brings me to this:

More Weakness is Likely

As I explained, BOIL is designed to track the Bloomberg Natural Gas Subindex such that a 1% daily rise in the index results in a 2% daily rise in BOIL.

I examined returns for the Bloomberg Gas Subindex since the end of 2002 over three distinct time periods — the period from year-end through February 9th of the following year, the period from February 9th through to April 5th and the period from February 9th through June 5th:

The first column shows the calendar year in question with each row relating to a particular year.

The second column shows the return for the Gas Subindex from year-end to February 9th; for example, for the first row labeled “2003” the second column shows the Bloomberg Natural Gas Subindex rallied 28.8% from the end of 2002 through the close on Friday, February 7, 2003 (the 9th was a Sunday that year).

The third column then shows the subsequent change in the price of the Subindex from February 9th through to April 5th (for weekends I used the last weekday close). The row labeled 2003 shows the index fell 10.5% from Friday February 7th, 2003 through Friday April 4th, 2003.

The fourth column shows the change in the Bloomberg Natural Gas Subindex from February 9th through to June 5th each year. For 2003, natural gas prices rallied 16.3% in total from February 9th to June 5th.

The final columns labeled “Signal Years” show the return for the Bloomberg Natural Gas Subindex in years where gas prices were weak to start the year. Specifically, this column only pertains to years when gas prices fell more than 11.4% between the beginning of the year and February 9th.

I didn’t choose 11.4% arbitrarily. Over the last 22 years, gas prices have fallen an average of 5.7% between the start of the year and February 9th, so this represents twice the average decline over the entire sample.

In years where that condition holds true, this column reports the return from February 9th through April 5th. The final column then represents the total change in the Natural Gas Subindex from February 9th through June 5th.

In 7 of the past 21 years, not including 2024, gas prices have fallen more than 10% to start the year on this basis. And in all those years, the Bloomberg Natural Gas Subindex declined further between February 9th and April 5th. Also, in all those years, the subindex saw a negative return between February 9th and June 5th.

The average decline in these 7 years was about 15.3% from February to April and 19.7% from February to early June.

In other words, in years where gas prices see a weak start to the year, there’s rarely a significant price recovery through the spring and early summer months.

I must warn you I am not a proponent of making investment or trading decisions based on complex, narrowly defined seasonal patterns. The danger is what behavioral economists called the “clustering illusion” – a psychological bias to look for patterns, relationships and meaning even in random data where no such patterns exist.

I only consider seasonals if I can discern some reasonable fundamental or economic justification for the pattern in the data.

In this case, however, these dates were not chosen at random, and I believe the seasonal pattern is not an exercise in data mining:

Seasonal Gas Demand and the Roll

In essence, this simple signal works because when gas prices decline sharply to start the year that signals deteriorating fundamentals at a time of year when gas demand is highest.

And, if gas fundamentals are weak to start the year, they’re unlikely to improve through the spring as gas demand seasonally weakens and storage levels tend to build – with few upside catalysts for gas at this time of year, prices have a tendency to weaken further.

As I mentioned earlier, the key season for natural gas prices is winter, not the summer months.

Take a look:

This chart shows data from the Energy Information Administration (EIA) on monthly US natural gas consumption in billions of cubic feet (the monthly total not bcf/day) from November 2012 through November 2023, the last month for which the EIA released monthly gas data.

The long black arrows I’ve drawn on my chart show the winter spikes in US natural gas demand related to demand for heat, while the shorter red arrows indicate the summer spikes in natural gas consumption related to electricity generation for summer cooling. There’s undoubtedly a pop in summer gas demand; however, it pales in comparison to heating demand.

In addition to this annual seasonal pattern, you’ll also notice overall US natural gas demand has been rising since roughly 2017 – there’s a pattern of higher highs and higher lows in the chart. Part of that is due to increased use of gas in electricity generation, replacing retired coal-fired power facilities.

According to EIA, coal accounted for some 38.9% of total US net electricity generation at utility-scale facilities in 2013 while natural gas chipped in 27.7%. By 2022 coal’s share of US power generation on this basis had fallen to 19.7% while gas had grown to 39.9%.

Increased use of gas in power generation does benefit both seasonal summer and winter demand to the extent some consumers use electricity for heat as well as cooling. However, take a quick glance and you’ll see that despite the rise in gas demand and the increased importance of the fuel to US power generation, there has been little or no change in the annual seasonal pattern of demand.

In short, winter weather remains the key swing factor every year.

So, consider this chart:

The National Oceanic and Atmospheric Administration (NOAA) reports heating degree days (HDDs) each season, which compare daily mean temperatures to a 65-degree Fahrenheit baseline (that’s about 18.3 degrees Celsius if you favor the metric system). For each degree Fahrenheit below that baseline, NOAA records 1 heating degree day for that date.

The data I present in this chart shows HDDs on a cumulative basis each season for the entire US, weighted by population to reflect historical regional natural gas consumption. Specifically, I’m looking at cumulative HDDs from the start of the winter heating season in late autumn through the first Saturday in March – the higher the number, the colder the weather.

As you can see, 2013-14 and 2014-15 were colder-than-average years on this basis. Flip back to my last chart and you’ll see this corresponds to larger-than-average winter spikes in US gas demand.

Similarly, 2015-16 and 2016-17 were warmer-than-average winters and the consequent spikes in winter gas demand were lower than in prior years.

Let’s put this in a different way.

Based on the EIA natural gas consumption data I described earlier, the two peak-demand months of January and February have accounted for an average of 20.8% of total annual gas demand in the years 2020-2022 inclusive. There are a total of 59 days in these two months, so January and February account for only 16.2% of calendar days in the year but almost 21% of average demand.

In contrast, the months of July-August include 62 calendar days (almost 17% of the calendar year) but have accounted for an average of only 15.9% of annual gas demand over the same three-year period.

Of course, patterns in gas demand and production also impact the storage cycle:

Withdrawals and Injections

The EIA’s weekly natural gas storage report is the most widely watched piece of data regarding the US natural gas market.

In it, the EIA reports the amount of natural gas held in underground storage across the Lower 48 US states and broken down by region.

Here’s a look at weekly gas storage since the beginning of last year as well as the 5-year seasonal average for each week over the same period:

First up, the orange line represents the 5-year trailing seasonal average. As I’ve labeled, US gas storage typically reaches its annual lows around the end of March each year and peaks in early November.

This primarily reflects the seasonal demand trends I just outlined. During heating season from November through March, US demand for gas is elevated, so the US consumes more gas than it produces, drawing on storage to fill that gap.

In addition, in some years, winter weather is so severe it interrupts gas production in certain states. For example, in mid-January this year, extreme cold resulted in well freeze-offs – shut-in production – causing a short-term drop in US natural gas production of around 10.6 bcf/day.

This period is known as withdrawal season.

On the flip side of the equation, US gas demand declines seasonally after heating season and, as I illustrated earlier, the summertime bump in demand is far smaller than the winter surge.

In this period of the year, the US generally produces more gas than it consumes, resulting in building supplies between (roughly) late March and the first or second week of November. That’s known as injection season.

So, one of the simplest and most powerful indicators of supply and demand conditions in the US gas market is to compare current storage levels, and the recent trend in storage, to the seasonal norm.

For example, US gas prices – the front month and the futures curve I showed you earlier — have generally fallen since the end of October last year. Look at my chart above and you’ll see US gas storage ended injection season last November well above the 5-year seasonal average level (excess supply).

Moreover, since the start of withdrawal season, US natural gas storage fell at a slower-than-usual pace through the end of last year.

Specifically, based on the data I presented in my chart above, US gas storage was some 165 bcf above the 5-year average as of November 3, 2023, rising to a whopping 387 bcf above average as of December 29th. Per EIA, US gas storage was running at 7-year seasonal highs in early January 2024.

Of course, as you might expect, the widely publicized Arctic blast in the middle of January 2024 helped drive elevated heating demand and draw down storage at a much faster-than-normal rate. Based on the data in my chart above, the storage excess relative to the 5-year average had shrunk to around 120 bcf above the 5-year seasonal average as of the end of January.

All these wild weather swings are too little, too late to save the US natural gas market this year.

While we saw cold weather in mid-January, the 2023-24 heating season so far has been about average:

This shows the cumulative natural gas weighted HDDs for the US since the end of September compared to the 5-year average. You can clearly see how the line representing the current 2023-24 season (orange line) pulled above the 5-year seasonal average (blue line) in mid-January.

However, a warm end to the month pushed the cumulative HDDs back down closer to average, balancing out the prior weeks’ cold.

That’s why, in the most recent gas storage report released last Thursday, EIA reports that US natural gas storage fell only 75 bcf in the most recent week, well below the normal decline for this time of year. You can see that US gas storage is now close to 240 bcf above the 5-year seasonal average once again.

And there’s no relief in sight:

At the end of January, NOAA predicted high probability of above-average temperatures through the current month of February and across most of the Lower-48 US states including key gas consuming regions of the Upper Midwest.

My point is simple – with US gas storage still well above seasonal normal levels, we’re running out of heating season this year. Specifically, the gas market is running out of high-demand weeks that could bring US gas storage closer to the 5-year average before heating season comes to an end next month.

Back to the Seasonal Signal

So, that brings me back to the logic behind the seasonal pattern I outlined earlier.

Specifically, when the Bloomberg Natural Gas Subindex falls more than 11.4% year-to-date through February 9th, natural gas prices tend to fall further through the late spring (through early June).

The fundamental logic of this seasonal pattern reflects the crucial importance of the months of January and February each year.

These are the two months where a cold snap can have an outsized impact on gas demand and those all-important weekly storage numbers from EIA.

Simply put, by early February each year, markets have a good handle on demand conditions through winter heating season. Obviously, by February 5th each year we already know about demand conditions and HDDs through the preceding three months (November through January).

In addition, weather forecasts for likely temperatures over the next few weeks are more accurate than longer-term seasonal forecasts for winter heating demand NOAA put out before the start of the season back in November.

Weather forecasts are always subject to error, and there can be surprise cold snaps well into March. However, by early February each year, market participants have a better idea of likely HDDs and natural gas demand through the peak weeks of heating season.

So, when natural gas prices fall significantly into early February that usually reflects weak heating season demand and excess gas in storage. For example, consider the 2016-2017 storage situation:

In early 2016, US gas storage remained consistently above the trailing 5-year seasonal average; as I explained earlier, the winter of 2015-16 was far warmer than 2013-14 and 2014-15, leading to weaker winter gas demand.

The table I posted earlier in this article shows that the Bloomberg US Natural Gas Subindex weakened 11.4% between the end of 2015 and February 9, 2016. And with storage well above normal heading into a period of weak demand from March-May, gas prices couldn’t catch a bid – the Bloomberg Subindex fell an additional 11.9% from February 9th through April 5th and was flat (down 0.9%) from February 9th through June 5th.

The following winter of 2016-17 was even warmer – the warmest of any season from 2013 through 2023 on the HDD basis I outlined earlier. So, while storage was close to the 5-year average in early 2017, by early February it was clear the withdrawal season would end with storage well above average.

The Bloomberg Natural Gas Subindex fell 14.7% from the end of 2016 through February 9, 2017 amid a worsening storage picture. Then, gas prices fell -0.6% from February 9th to April 5th, 2017, and a total of 13.2% from February 9 to June 5, 2017.

As I previewed earlier, when gas prices are weak even amid the peak demand period early in the year, prices usually continue to fall into June. That’s because once heating season is over, there are few obvious upside catalysts for gas until summer demand picks up in July and August.

And with the Bloomberg Natural Gas Subindex plummeting more than 20% from the end of 2023 to February 9th, this signal has tripped once again – given just how weak gas fundamentals are right now, we’re unlikely to see a major rebound over the next 3 to 4 months.

And that brings me to this:

Natural Gas Problems Compounded

As I said, BOIL is designed to track twice the daily performance of the Bloomberg Natural Gas Subindex. That means if the Subindex is up 1% in a single trading day, BOIL aims to return 2% on that trading day.

Bloomberg outlines its methodology for pricing the Natural Gas Subindex in its Index Methodology report posted here.

Here’s the futures roll scheduled for BOIL:

The first column in this table shows the calendar month of the year and the second shows the natural gas futures contract tracked by BOIL in that month.

So, for example, in January 2024 BOIL tracked the March 2024 natural gas futures contract for the entire month. This month however, BOIL began selling its exposure to the March 2024 futures and rolling into the May 2024 futures on the fifth business day of February. The index transfers 20% of its exposure from the March 2024 to the May 2024 contract per business day, starting on the fifth business day of February, and continuing until its 100% tracking the May futures.

Next month, BOIL will continue tracking the May 2024 futures for the entire month. Then, starting on the fifth business day in April (Friday, April 5th), BOIL will sell May 2024 futures, buying the July 2024 futures, and transferring 20% of its exposure to July each business day until the roll is complete.

There are two crucial points about the way BOIL rolls its futures market exposure.

First, when you own the May 2024 futures contract, the only fundamentals that can move prices are changes in the supply and demand outlook for gas between now and May. So, any forecasts for hot summer weather in July and August won’t have any fundamental impact on the price of gas for delivery in the month of May.

Further, expectations for the new Golden Pass LNG export terminal in Texas to begin exporting significant quantities of US natural gas by early 2025 may well be a bullish factor for gas prices longer-term. The US LNG export boom out through at least 2027 is a key plank in my bullish thesis for gas producer Southwestern Energy (SWN), outlined here on Seeking Alpha back on December 6th.

However, Golden Pass will do nothing to change the supply-demand balance for US natural gas in May 2024 since it’s not likely to start pulling additional gas demand until early next year.

So, over the next two trading days, BOIL will complete the roll from March 2024 gas futures to the May contract. The price of May 2024 futures will likely vary based on the outlook for natural gas storage around the end of withdrawal season in late March and in the early stages of injection season.

As I outlined at length earlier, the news here isn’t good – storage is around 240 bcf above normal for this time of the year, peak heating demand season is nearing an end and NOAA weather forecasts show above-average temperatures likely through the end of heating season.

This is yet another reason I monitor the performance of the Bloomberg Natural Gas Subindex year-to-date through February 9th – this date is significant because it falls around the time the roll in BOIL’s exposure from the March to May futures contract is underway.

That’s why it’s also crucial to watch the index roll starting in early April. Since BOIL rolls from the May contract, a shoulder season for natural gas demand, to the July contract, expectations for summer cooling season can start to have some impact on the price.

However, if you flip back to my table early in this article you’ll see I report the historical performance of the Bloomberg Natural Gas Subindex over two periods following weak starts to the calendar year – from February 9th to April 5th and from February 9th to June 5th, around the time BOIL begins the roll from the July to the September futures.

In years where natural gas prices are weak early in the year, that weakness tends to persist through early June, through the roll into the July contract in April. Indeed, in 5 of the 7 “signal” years since 2003, as I outlined earlier, weakness in gas actually worsened between April 5th and June 5th.

I believe the reason is related to the seasonal demand cycle I outlined earlier. Simply put, the summer cooling demand pull is much smaller than the winter heating season, so when fundamentals for gas are weak through the end of the heating season in late March, it’s tough for that situation to improve enough to change the supply-demand balance by June 5th when the Natural Gas Subindex begins the roll to the next futures contract.

There are two more headwinds for BOIL right now.

First, the calendar roll.

There can be significant cost associated with rolling from one futures contract to the next over time when the market is in a state known as contango. Contango simply means the price of gas for delivery in future months is higher than the price of gas today.

As I showed you earlier, that’s the case for gas prices over the next few months. Bloomberg reports that on Friday February 9th the March 2024 contract closed at around $1.85/MMBtu, the May 2024 contract closed at $1.99/MMBtu and July 2024 at $2.35/MMBtu.

Part of the reason gas prices are higher for future delivery is that it costs money to store gas for delivery in the future – the higher futures price compensates for the cost of that storage.

However, over time, if fundamentals for gas don’t change much, there’s a tendency for the price of futures to gradually fall more in line with the current front-month or spot price.

In other words, if there’s little change in the supply demand outlook for natural gas between now and early April, one might expect the price of those May 2024 futures to decline to around $1.85/MMBtu, the current price of the March 2024 futures. Again, part of that is due to the erosion of the storage premium factored into May futures over time. In addition, as I explained, there are few obvious weather catalysts for gas between now and May 2024, historically one of the weaker months for gas demand.

So, with the market in contango through this summer, every two months, BOIL is essentially selling a low-priced front-month contract and buying a higher priced contract for delivery further into the future. Each month, this roll might only cost BOIL a few cents; however, over time these roll costs add up.

Second is the compounding risk.

Consider two hypothetical assets — Asset A and Asset B — as well as leveraged ETFs that track those two assets, designed to rise 2% for every 1% daily increase in the underlying:

As you can see, the price of Asset A rises 2% each day for 10 trading days, a total of 21.9%. The price of Leveraged Asset A, therefore, rises 4% each trading day over the same period, for a return of 48% over 10 days.

So, while leveraged asset A tracks the daily change in Asset A perfectly – rising twice the daily percentage increase – the return over 10 days is a little over 48%. Of course, that’s more than double the 21.9% increase in Asset A over 10 days.

Now, consider Asset B. As you can see, this asset rises a bit more over 10 trading days than Asset A, for a total return of 22.6%.

However, Asset B is more volatile – the day-to-day swings in price are more jagged. Now, Leveraged Asset B also tacks daily changes in Asset B perfectly, rising or falling by twice the daily percentage change in the underlying each trading day.

However, after 10 trading days Leveraged Asset B is up only 38.2%. That’s far less than twice the 22.6% gain in Asset B over those 10 trading days. And it’s less than the 48% gain in Leveraged Asset A despite the fact Asset B was up more than Asset A over the same 10 days.

Two points:

First, leveraged ETFs like BOIL track twice the daily price change in the underlying asset, in this case the Bloomberg Natural Gas Subindex. Returns over longer holding periods can deviate significantly from twice the performance of the underlying. The longer the holding period, the larger the potential deviation in performance as the leveraged daily swings compound.

Second, the more volatile the underlying, the larger the potential deviation in performance over time.

This second point, in particular, is problematic for natural gas, one of the most volatile commodities you’ll encounter:

This chart plots the 30-day trailing annualized volatility in the Bloomberg Natural Gas Subindex since February 2016. The average annualized volatility over this time period (orange line) is about 47% and, as you can see, natural gas price volatility recently jumped above that long-term average for the first time since last summer.

That’s not good news for BOIL. Indeed, even if the Bloomberg Natural Gas Subindex does manage a modest recovery into June, increased volatility and the compounding issue over time could conspire to push down the price of BOIL.

This has happened before. Per Bloomberg, the Natural Gas Subindex rose from 85.04 on April 14, 2023 to 86.82 on August 30, 2023, a gain of 2.1% over about 4.5 months. Over the same time frame, Bloomberg indicates the price of BOIL fell 3.2%.

Conclusions

With front-month natural gas futures well below the psychologically important $2/MMBtu level it’s tempting to suppose buying BOIL, a leveraged ETF tracking the Bloomberg Natural gas Subindex, would be an advantageous way to profit from a recovery in gas prices from multi-year lows.

However, US gas storage levels are plentiful right now and the near-term supply-demand balance is bearish.

History shows that when natural gas prices fall significantly through January and into February – the heart of winter heating season – gas usually falls further through the spring and into the month of June.

Heightened volatility in gas prices, and contango in the gas futures curve, increase the downside risks for BOIL between now and early June when the ETF will roll to the September 2024 futures contract.

If you’re seeking exposure to the price of gas over the longer term, consider buying a high-quality gas producer like Southwestern Energy, which I profiled here on Seeking Alpha in early December or a name like Range Resources (RRC) profiled here last summer.

As I outlined in those articles, producers like SWN and RRC have hedges to protect cash flows from current weakness in gas prices and, longer, term are well-positioned to benefit from a rally in gas as new LNG export terminals come onstream. Buying a producer also avoids the compounding and calendar futures roll risks that plague BOIL.