The EIA announced that natural gas inventories rose by +20 BCF for the week of July 22-28 in its weekly natural gas storage report issued Thursday morning. The injection was 2 BCF larger than my +22 BCF projection and a bullish 24 BCF smaller than the 5-year average. The bullishness of the small inventory build was driven by the South Central Region which saw a -17 BCF weekly withdrawal, 16 BCF bullish versus the 5-year average -1 BCF draw, making the region responsible for 2/3rds of the nationwide departure from the 5-year average. It was the second straight week in which the region saw a net draw. Storage levels in the South Central Region are now down 95 BCF from the same week in 2016, the largest year-over-year deficit of any region. The consistently bullish reports coming out of this region this summer are likely due to increased feedgas demand to the Sabine Pass LNG liquefaction plant on the Texas/Louisiana border, which has more than doubled year-over-year to average around 2.2 BCF/day this summer. On the other hand, the East Region was the only region of the five to report a bearish injection, +25 BCF versus the 5-year average +17 BCF, the second straight week the region has reported a larger-than-normal injection despite a much smaller-than-average nationwide build. This is likely attributable to seasonally cool temperatures across the region, but also to record production coming out of the Marcellus and Utica shales in the Appalachians.
With the +22 BCF nationwide injection, total natural gas inventories rose to 3,010 BCF, with storage crossing the 3,000 BCF mark 2 weeks ahead of schedule. Nonetheless, the storage surplus versus the 5-year average narrowed to +87 BCF or +3%, the first time it has been under +100 BCF since February 10, 2017. 2 storage regions--the Northeast and Pacific--are currently at deficits versus their respective 5-year averages. The year-over-year storage deficit, meanwhile, narrowed by 23 BCF thanks to last summer's remarkable -3 BCF nationwide storage withdrawal to a still-solid -279 BCF or -8%. All 5 storage regions are at comfortable year-over-year deficits with each exceeding 5%.
Current regional and national natural inventory levels are summarized in Figure 1 below.
Figure 1: Summary of Thursday's EIA Storage Report For July 22-28 [Source: CelsiusEnergy.Net via EIA Data]
Despite generally seasonal temperatures this summer--especially compared to 2016--storage injections have been persistently bullish. Since natural gas inventories bottomed out on March 24, a total of +961 BCF has been added to storage during the first 18 weeks of the injection season. This is a bullish 308 BCF less than the 5-year average. A +961 BCF 18-week build would be the third smallest in the last 5 years behind only 2016's +820 BCF and 2012's +794 BCF, with the important caveat that each of those years had much higher withdrawal-season ending inventories. Interestingly, the +961 BCF build would also be the third smallest in the full 23 year period for which EIA storage data is available. In contrast, 2014 saw +1566 BCF injected in its first 19 weeks of the season, 63% higher than 2017 and the most of any year in the last 23 years. 5-Year and 23-year data is summarized below in Figure 2.
Figure 2: Injection-season to date storage statistics showing a well-below-average 18-week storage injection [Source: CelsiusEnergy.Net via EIA Data]
On Thursday afternoon, the EIA also released its weekly natural gas supply/demand data via PointLogic, which provided insight into supply/demand balance.
Total natural gas demand was nearly steady last week, falling less than 0.1 BCF to average 74.4 BCF/day, which trailed 2016 by -2.6 BCF/day. The majority of this deficit was attributable to powerburn which, while it rose by 0.5 BCF/day week-over-week to a new 2017 high of 36.7 BCF/day, stilled trailed 2016 by a large 2.9 BCF/day, due to both a tighter market last year as a result of (even) cheaper gas prices and hotter temperatures last year. Additional year-over-year losses were driven by residential/commercial demand. However, these weather-dependent losses were mitigated by strong gains in natural gas exports, particularly LNG feedgas demand which averaged 2.4 BCF/day last week, a new record high and 1.5 BCF/day higher year-over-year.
On the supply side, natural gas domestic production continued to rise, climbing another 0.2 BCF/day week-over-week to average 72.2 BCF/day, an ugly 1.3 BCF/day higher than the same week in 2016. Since bottoming at under 70 BCF/day in late April, natural gas production has climbed by a bearish 2.5 BCF/day over the past 3 months. The growth in natural gas production versus 2016 is shown in Figure 3 below.
Figure 3: Natural gas production over the past year compared with year-ago levels [Source: CelsiusEnergy.Net via EIA Data]
Based on data through Wednesday, August 2, it appears likely that natural gas production will continue to climb into next week, possibly reaching 72.7 BCF/day. And with further takeaway capacity from the Northeast shale regions due to come online via the Rover pipeline in coming months, it would not be surprising to see production continue to rise in coming weeks.
Fortunately, the effect of rising domestic production has been mitigated on the supply side due to a sharp decline in Canadian imports, which stood at 6.0 BCF/day down 0.1 BCF/day week-over-week and down a steep -1.5 BCF/day from 2016. As a result, total natural gas supply was up less than 0.1 BCF/day week-over-week at 78.2 BCF/day and was down -0.2 BCF/day from a year ago.
There is no doubt that natural gas demand is highly influenced by changes in temperature which is responsible for the commodity's increased volatility relative to others in the energy sector. However, I argue that, with a few exceptions such as the polar vortex winter of 2013-2014 and the summer-in-spring 2012, variations in temperature do not drive large, long-term changes in the natural gas storage picture. Instead, it is the supply and demand variables driven by changes in the economics of natural gas versus competitor fuels that cause smaller week-to-week but long-lasting incremental changes that truly shape natural gas supply/demand balance. Temperature-independent supply/demand balance essentially "unmasks" the true fundamental health of the natural gas sector as, given enough time, temperatures will normalize. These temperature-independent variables include domestic production, Canadian imports, and LNG imports on the supply side and LNG exports and exports to Mexico on the demand side. Declines in temperature-independent supply variables and gains in temperature-independent demand variables equate to a tighter market and vice-versa. I view this temperature-independent supply/demand balance as one of the best metrics for estimating the overall health of the natural gas market.
For the week of July 22-28, I calculate that temperature-independent supply/demand balance stood at +1.8 BCF/tight versus 2016, +0.2 BCF/day tighter week-over-week. As mentioned above, temperature independent market tightness last week was driven by large year-over-year declines Canadian imports (-1.5 BCF/day) and gains in LNG imports (+1.5 BCF/day) with a bearish gain in production (+1.3 BCF/day) countering this tightness. Mexican exports and LNG imports are both essentially flat year-over-year. A summary of temperature-independent supply/demand data is shown in Figure 4 below.
Figure 4: Temperature-Independent Natural Gas Supply/Demand Balance Versus 2016 showing ongoing modest market tightness [Source: CelsiusEnergy.Net via EIA Data]
As the Figure shows, natural gas supply/demand tightness has decreased from January-April when it averaged 4-6 BCF/day tight versus 2016, largely thanks to gains in production. However, year-over-year market tightness has increased slightly each of the past 3 weeks since dropping to near +1.0 BCF/day tight in early July. As a result, I view that natural gas market as modestly supportive on a long-term basis. It is worth mentioning that this temperature-independent supply/demand balance calculation does not include powerburn as that variable, while partially dependent on the economics between natural gas and coal or other competitor fuels, is also dependent on temperature, particularly during the summer, to drive cooling demand. As mentioned above, 2017 powerburn has consistently trailed 2016 due to the combination of cheaper gas and hotter summertime weather.
Overall, I view Thursday's storage injection and accompanying supply/demand data as modestly bullish. The natural gas storage surplus has fallen to 5-month lows despite less-than-supportive temperatures and temperature-independent natural-gas supply/demand balance remains tight. Any other year in the last 5 besides 2016, such a pattern would have likely resulted in a near-average +40 BCF to +45 BCF storage injection, but instead the storage surplus contracted by nearly 25 BCF. While temperature-independent supply/demand balance will probably loosen some in coming months as new supply comes online, this trend will be countered in part by gains in LNG feedgas as Sabine Pass' Train 4 comes online. Plus, the market was so tight in 2016 that it doesn't necessarily need to maintain year-over-year tightness to continue to contract versus the 5-year average. I fully expect this surplus to flip to a deficit by the end of September or early October. As a result, I believe that it is a foregone conclusion that inventories will finish under 4,000 BCF for the first time in 3 years--I am projecting 3780 BCF at this time--and likely the second lowest in the last 5 years. Therefore, natural gas could be more susceptible to a price spike should winter get off to an early start compared to recent years when 4000+ BCF inventories provided a large storage buffer that made any concerns about a supply crisis largely moot.
At least right now, natural gas investors don't seem to agree with my assessment. After falling to 5-month lows early this week on fears of cooler-than-average temperatures throughout August, the commodity erased a 1% rebound pre-report to finish Thursday down 1 cent or 0.4% right on the day and settle near the recent lows at $2.80/MMBTU. The 1x ETF United States Gas Fund (UNG) was appropriately down 0.3%. The leveraged ETFs, on the other hand, underperformed to the downside thanks to some rebalancing issues over the past few days due to moves in the Futures price after the 2:30 PM close of commodities trading. The 2x leveraged ProShares Gas Fund (BOIL) fell -1.4% versus a predicted -0.6% drop while the 3x leveraged VelocityShares product (UGAZ) fell -2.1% versus a predicted -0.9% drop. Its inverse cousin DGAZ similarly outperformed, climbing just under +2.1%.
Based on a current and projected fundamental analysis, I stubbornly continue to feel that natural gas has overextended to the downside due to fears about the near-term temperature pattern. According to my Fair Price model, which compares current and projected inventories versus historical inventories and prices over the past 5 years, natural gas is trading at a very steep 13.5% undervaluation based on current inventories alone with a Fair Price of $3.23/MMBTU versus the observed Futures price of $2.80/MMBTU. Because of near-term storage projections and the longer-term market tightness, the 8-month average undervaluation grows to 14.4%. This data is shown below in Figure 5.
Figure 5: Natural gas undervaluation versus Fair Price based on projected inventories for the next 8 months[Source: CelsiusEnergy.Net via EIA Data]
As a simple back-of-the envelope illustration, at $2.80/MMBTU, natural gas prices are at their lowest since March 1. At that time, natural gas inventories were +363 BCF versus the 5-year average. As of this morning, at an identical price of $2.80/MMBTU, I estimate that the surplus is +75 BCF. Thus, prices are about the same despite a 280 BCF improvement in the storage surplus. This is rather paradoxical. It would be one thing if the surplus was poised to rise sharply in the upcoming weeks. However, the opposite is true as I expect the surplus to continue steadily contracting or hold steady.
For this reason, my price target for the end of the injection season in late October or early November is rather conservatively set at $3.15/MMBTU which would be a 12.5% rally from current levels and would bring natural gas to just under its average Fair Price. I would not be surprised to see some further near-term weakness in natural gas price because, based on the latest model guidance, August does look honest-to-goodness cold, unseasonably so. Thereafter, however, I expect prices to rebound. And should August turn out warmer than expected, I would not be surprised to see prices turn even higher. I remain long natural gas based on these projections via my preferred long-term trading strategy, short DGAZ to gain long exposure and capitalize on leverage-induced decay.
Looking ahead to next week, the EIA will release its Storage Report for July 29-August 4 next Thursday, August 10 at 10:30 AM EDT. I am projecting a preliminary +40 BCF storage injection, which would be 14 BCF smaller than the 5-year average +54 BCF. It would be the third smallest injection for this period in the last 5 years, behind only 2012's +23 BCF and 2016's +24 BCF. It would, however, by 16 BCF larger year-over-year. Should a +40 BCF build verify, natural gas inventories would rise to 3050 BCF while the storage surplus versus the 5-year average would continue to retreat, falling to +73 BCF, the lowest since February 3.
Disclosure: I am/we are short DGAZ. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.