- We only expect a notable step-up in sequential production growth at the close of the injection season and start of heating season 2019-20 given expected gains tied to the 2.0 bcf/d Gulf Coast Express entering service. Notably, we have seen a considerable slowing of permitting activity across major gas basins like Appalachia and Haynesville. For Appalachia, preliminary Q1 19 data (which is subject to upward revision) show a marked slowdown in the rate of well completions too. The current forward curve is also not doing much to incentivise gas-directed drilling either. Our balances include average injection season production growth of 7.6 bcf/d y/y, slowing to near 4.5 bcf/d y/y over the heating season.
- To prevent storage overfill, gas prices have declined to keep demand in the power sector close to flat season-to-date. Gas has taken further market share from coal in the generating stack, with gas climbing to a nearly 60% share on our estimates, from 51% a year ago. Lower total power load has reduced the overall call on coal and gas by more than 6 bcf/d. The increased share for gas implies the gas equivalent of a more than 5 bcf/d reduction in coal generation. At least half of that is attributable to switching y/y from coal to gas due to the economic incentive to do so, with another 1 bcf/d due to structural factors such as coal retirements of more than 4.4 GW y/y.
- The Henry Hub-TTF arb for Aug-19 contracts is narrowly open on paper once more, even as a massive y/y storage surplus in Europe remains. However, our reference case assumes no US LNG trains will be shut in as cargoes will likely have been hedged previously and the daily freight rate may be a sunk cost for many industry players. However, that does not rule out the possibility of lower feedgas flows into terminals, thereby reducing monthly cargoes sent out into the market. Of note, our October balances assume one generic US LNG train will be out of service for maintenance for a month.
- The bias is still to the downside on demand growth given the strides already made into the power dispatch queue and the high gas burn print we envision for July. For Mexico, LNG cargoes, which have already been tendered, into Altamira through late July leave LNG intake flat y/y for most of the peak cooling season, suggesting that aside from linefill, testing and minor flows on Sur de Texas-Tuxpan, there will not be a marked ramp-up on the pipe until late July and August as that LNG backfills demand. While new Gulf ethane crackers are taking in feed, none of the three expected online this quarter are world-scale and they will only push up incremental demand narrowly. In addition, the recent consensus analyst estimate misses versus EIA for weekly storage injection suggests that demand is underperforming.
Storage and price outlook
- Our storage carryout of 3.72 tcf is still within 60 bcf of the estimates that we showed in March and nearly flat to the previous two months. However, that glosses over the 0.9 bcf/d average upward adjustment to August-October power demand on lower price. A notably lower carryout is dependent on hotter weather, but given how much work price has already done, hot weather can only do so much heavy lifting. In that vein, we see little upside to our gas burn in power forecast for July. A period of 5% hotter-than-normal temperatures would add 0.6 bcf/d to burn on average in both August and September, knocking only 35 bcf off our forecast end-October carryout. A decidedly more bullish hot weather scenario—10% warmer than the 10-year normal in those two months—would knock off close to 60 bcf from the carryout.
- Our price outlook for the remainder of the injection season is $2.25/mmbtu on the limited demand growth we see and our forecast end-October storage carryout is 3.72 tcf. That translates to an end-March 2020 carryout near 1.6 tcf, which should pressure winter prices to $2.52/mmbtu.
- Of note, there is currently a small cash versus futures gap with Jul-19 to Sep-19 contracts and cash price so far in June averaging in the low $2.40s/mmbtu. CME options open interest points to a market that has evolved over the past month toward seeking more downside protection than upside and CFTC open interest data show a very short bias.