Today’s report (week ended 5 January): EIA: -359 bcf, EA: -341 bcf
- Today’s reported withdrawal was much higher than the 330 bcf consensus estimate. The gap between the reported figure and the forecasts is a function of the unprecedented nature of the storage draw—a massive 10 bcf/d higher than the previous record—and the resulting stress on both flow- and supply-demand based models.
- While we settled on a 341 bcf withdrawal for this week, our forecasts as of last week, amid the ‘snow bomb’, had hovered around 350 bcf. With temperatures at such bitterly cold levels, the gas-heating intensity (GHI) (mmcf/d usage per GWHDD) begins to break down as increases in GHI begin to level off at such extreme cold. Our miss was consequently concentrated in the res-com sector, though we did upwardly adjust power by 0.3 bcf/d following this morning’s report. Our forecast of -341 bcf was at the high end of analyst estimates in the Bloomberg survey (the highest estimate was -345 bcf), yet still off by 2.6 bcf/d from the reported figure. This underscores the difficulty in assessing a record-breaker of such magnitude.
Next Thursday’s report (week ending 12 January): EA preliminary: -189 bcf
- Our forecast for res-com demand has been revised lower versus Tuesday’s estimate, by 0.3 bcf/d, and we have also revised power down by 0.8 bcf/d.
- While our Permian sample has indicated a recovery to pre-freeze-off baseline output, forecast low temperatures beginning this evening through the middle of next week once more put the basin on our watch list for potential losses.
Storage limbo: How low can it go
Today the EIA reported a record-breaking withdrawal of 359 bcf, blowing the previous record from the 2013-14 Polar Vortex winter out of the water by 71 bcf (10 bcf/d). Record-breaking is not a term to casually throw around. However, it is entirely appropriate for so many aspects of the market last week beyond the withdrawal, particularly with such high consumption in the Northeast and prices at Transco Zone 6 trading as high as 175 $/mmbtu. The same cannot be said for recent price action. The market barely lost its stride, with the prompt month intra-day high tapping out at just under 3.1 $/mmbtu in the midst of ferocious cold. Cash Henry Hub prices saw a short-lived response, averaging 4.40 $/mmbtu for the reference week, peaking at 6.88 $/mmbtu, a high not seen since the Polar Vortex winter. That the storage draw was 10 bcf/d above the previous record underscores just how abnormally tight balances were in the week ending 5 January, even if price did not send that signal.
With comparisons continuing to be drawn against the Polar Vortex in the short term, an appropriate question would be what has tempered the market’s reaction and what about the fundamental backdrop is so different than the cold of that fabled winter.
The most simplistic answer is of course to place all culpability on a pre-cold baseline of production growth nearing 6.5 bcf/d y/y, as well as a summer outlook for growth easily approaching 7 bcf/d. This supply-heavy backdrop translates into an outlook that does not involve much risk of supply scarcity for the remainder of this heating season. From a more short-term perspective, prices were not supported by the current thaw (which features near-normal temperatures) following the cold snap. As a result, for the week ending 12 January, the expected storage withdrawal has been continuously revised downwards, from an estimated 220 bcf in last week’s Panorama to 189 bcf as of this morning. In addition, price momentum behind cold tends to be stronger at the start of the heating season.
To focus only on supply growth and the thaw is to risk ignoring remaining weather-related risks to end the heating season. With the most recent weather forecast revisions now calling for another cold spell to follow the thaw, the pull from storage for the week ending 19 January is currently estimated at -278 bcf, another very strong withdrawal, on GWHDDs which are both far colder than the 10-year normal and the year-ago level. While this is contingent on the accuracy of those weather forecasts, the market is likely already pricing in that probability of another outsize pull from working gas stocks. Beyond this near-term withdrawal activity, the ultimate question is how low can storage go? A plausible, yet very bullish cold scenario for February and March is helpful for understanding any remaining weather risks and, perhaps more importantly, why price action has been subdued in the wake of such cold.
Our reference case sees an end-March storage carryout of 1.38 tcf, which assumes 10-year normal weather in February and March. To construct a low case end-March storage scenario, we assume 10% colder-than-normal weather in February and March. By comparison, February and March of 2014 were more than 15% colder than the 10-year average. In other words, while the 10% colder scenario is cold, we are not pushing the envelope on extreme temperatures either.
The 2013-14 Polar Vortex winter and this current heating season both started with stocks near 3.8 tcf. But, as today’s EIA report revealed, December ended with stocks some 100 bcf higher. Based on current weather forecasts, end-January inventories are still slated to be 75-80 bcf higher than in 2014. With only February and March left to work off those excess stocks, that equates to 1.3 bcf/d of overhang.
Given that the typical peak of GWHDDs is in January, the opportunity for growing demand on weather starts from a lower baseline of heating demand in February and March than it would in December or January. The uplift from the res-com sector under 10% colder-than-normal weather (vs the 10-year average) is an additional 200 bcf (3.3 bcf/d) of demand for those two months combined, according to our balances. For industrial heating loads, an additional 0.5-0.7 bcf/d of uplift is possible on such cold, which amounts to 35 bcf or so for the two-month period. Effectively then, another 235 bcf of demand would come from heating needs. Holding all other components of balances flat, that takes end-March storage down to approximately 1.15 tcf.
Of course, such cold makes another bout of supply freeze-offs likely. The geographical dispersion of cold will ultimately be the arbiter of how much supply will be lost. If cold on this level is concentrated in the eastern third of the country, the Permian, Bakken and other areas in the Midcon might not see the same impacts as seen earlier this month. But, for the sake of stacking the cards in favour of the lowest possible carryout, we assume total heating season production losses will be on the order of 100 bcf, near previous highs. With cumulative losses to date, that figure for the remainder of the heating season amounts to 60 bcf. Such events would take storage down to 1.09 tcf.
Outside of the biggest line items of heating demand in the res-com and industrial sectors and winter wellhead production disruptions, other components of supply-demand balances would also be affected by this notional 10% colder-than-normal scenario. Extra supply would enter our balances due to Northeast LNG sendout and higher net Canadian exports to the US. Such an increase in incremental sources of supply would easily push the end-March carry-out back toward 1.15 tcf. On the demand side, weather-aided uplift in the power sector should be expected in a 10% colder-than-normal scenario, with some trade-off between higher total loads and higher cash gas prices. But, with all the moving pieces in this cold scenario, it will take a very severe event (i.e. far more than 10% colder-than-average) to move the needle materially below an end-March carryout of 1.1 tcf. While such a figure certainly is below our current reference case concentrated in the 1.35-1.40 tcf zone, it is still well-above the 837 bcf end-of-season close in March 2014. Consequently, the seeming market complacency on such a record event is understandable as a very concentrated cold period leading to an extended run of high weekly withdrawals would be necessary to make storage scarcity a real concern come the end of March.
For the next two reports, we still expect hefty storage draws. For the week ending 12 January, we are forecasting a draw of -189 bcf, which, although sizeable, shows the tremendous step-down in demand compared to the week in progress. As mentioned earlier, for the week ending 19 January, we are expecting another large-scale draw of -278 bcf (versus -137 in 2017), vying for the second largest pull of all time. But, as the low end-March storage scenario suggests, mere normal weather will not lead to precarious storage positioning to close the season. As such, while if forecast cold does hold for the week ending 19 January, another bout of limited cash strength is likely, the prompt contract strength of today (+0.15 $/mmbtu as of midday New York time) is probably fleeting.