Today’s report (week ended 23 Nov): EIA: -59 bcf, EA: -66 bcf
- There was a wide range of estimates for today’s print, with consensus calling for a withdrawal in the mid/high 70s. Consensus, which missed the mark by nearly 2.5 bcf/d, seemed to fail to account for low demand due to Thanksgiving as well as some reversal in line pack from last week’s triple-digit inaugural withdrawal for this heating season. We estimated that the industrial sector ceded 1 bcf/d w/w due to the holiday impact. To true up to today’s number, we revised our w/w loss in the power sector to 5.3 bcf/d, vs 4.2 bcf/d in our Tuesday estimate.
Next Wednesday’s report (week ending 30 Nov): EA preliminary: -62 bcf
- Fairly similar GWHDDs w/w result in a res-com demand estimate only 0.4 bcf/d higher w/w. A 1.1 bcf/d surge in production is nearly offset by a 1 bcf/d w/w increase in power burn.
The F train
Yesterday’s pre-expiration spike in the Dec-18 contract, along with the rest of the winter contracts, added to concern about how much further the Jan-19 (F) contract has to go, or if it will just seemingly stop in its tracks—much like our beloved F train in New York. After significant early cold, the related shaving of inventories, and blowing through multiple levels of technical resistance on the way up, the market is searching for a sweet spot in pricing. That sweet spot needs to be one that ensures that enough power demand is being redirected away from gas and that deliverability concerns do not arise in the first half of Q1 19.
Price action this week was characterised by whipsaws on moves in weather forecasts, though yesterday’s up move was likely exaggerated by contract expiry. On Tuesday (27 November), the market showcased the impact of a then in-progress warming trend vs the abnormal cold last week together with new daily production records that were compounded by a forecast for a period of mild weather from mid-December to early January. Our view is that deliverability remains a potentially dire issue and that discernibly higher cash prices are still in the offing. Such concern was underscored by Dominion’s critical notice yesterday stating that injections and withdrawals at certain points on its system ‘are anticipated to be limited and or not available during the latter part of this winter season.’
Cash prices above $4.25/mmbtu and reaching as high as $4.80/mmbtu meant gas-fired generation ceded market share to coal through the past week. Gas burn in the power sector has been more than 3.8 bcf/d higher y/y in the past three weeks, but growth could have been nearly twice as high if coal had not ramped up as well. While both coal- and gas-fired generation ramped up in response to higher y/y load due to colder temperatures, thermal power was also needed to fill in for slightly lower nuclear output y/y. The early cold and low stocks for both gas and coal—with coal inventories at a four-year low on a days-of-forward cover basis heading into winter—could lead to even higher gas cash prices in December and January.
Our weekly balances are pointing to an end-November carryout of 2.98 tcf, more than 210 bcf lower than our balances tied to 10-year normal weather. The impact of November cold equates to 1.4 bcf/d of additional call on storage throughout the heating season. With our projected end-December storage carryout just below 2.5 tcf, the market would end 2018 still well below where inventories stood in December 2013, the year of the Polar Vortex.
Mild weather could easily pad our current balances toward an end-December carryout of 2.6-2.7 tcf depending on the degree of that warmth. However, warmth in the next five weeks would have to be extreme for the market to sound the all clear. The deliverability cushion is still necessary to get market participants with physical exposure at least through the first half of February. At that point, assured of their ability to deliver gas from storage, utilities/local distribution companies (LDCs) and physical market players could rely less on physical market purchases, which are helping prop the cash market up at the moment, and instead begin to rely on storage. From a financial perspective, the weighted-average cost of gas (WACOG) in storage is significantly cheaper than physical purchases and should continue to be so through the end of winter. However, now meeting the obligation to serve will be the primary objective of utilities/LDCs.
Even if moderate weather comes over the next five weeks, the market would still have to withstand the possibility of significant cold for the rest of the heating season. Late December typically represents an inflection point for GWHDDs in the US as deep winter cold sets in. Using 28 December as a baseline, the last five years have seen an average of 30% higher GWHDDs in the two weeks immediately following that date than they have in the two weeks preceding it. Such a dramatic spike this year in late December would likely erode any deliverability cushion and eliminate any storage buffer that a warmer-than-usual December might accumulate. Importantly, that spike in GWHDDs tends to occur after the expiry of the January-delivery contract, so the February contract could rise in price if that typical late December spike in GWHDDs comes to fruition.
|Fig 1: US coal generation, aGW||Fig 2: Percentage growth in HDDs after 28 Dec.|
|Source: Bloomberg, Energy Aspects||Note: Comparison is between the two weeks preceding 28 December and the two subsequent weeks.
Source: Weatherbell, Energy Aspects