Today’s report (week ended 22 Mar): EIA: net change: -36 bcf, EA: -33 bcf
- To align with today’s print, we reduced our production gain by 0.1 bcf/d, to a net 0.1 bcf/d w/w rise. The rest of the change involved raising res-com heating demand.
Next Thursday’s report (week ending 29 Mar): EA preliminary: +15 bcf
- A 1 bcf/d surge in production receipts and a 1.3 bcf/d w/w decline in LNG feedgas swings balances to a net injection for the first time since early November.
With global gas prices looking incredibly bearish, May-19 and Jun-19 JKM-TTF spreads have fallen to below where we estimate they need to be to keep the netback for US cargoes to Asia higher than to Europe. Given the TTF-Henry Hub spread suggests cargoes will head to Europe, the concern is when European demand for gas will be exhausted to the point that the arb to Europe closes, risking shut-ins of US LNG supply. With our present set of global balances, US exports are not constrained and end-October US storage is forecast to hit 3.57 tcf. Even if the arb closes, US balances do not suggest a tank top situation, as actual lost cargoes are likely to be limited and we expect some of the gas not exported to be diverted to price-induced domestic coal-to-gas switching. Of course, that implies global price weakness bleeding into the US market.
As we pointed out in last week’s Panorama, offtakers at Cheniere’s facilities have an approximate 60-day notice period to declare if they will not lift cargoes. In essence, that would mean that a decision right now to not lift volumes for LNG export would be felt in June, which will soon become the front-month JKM contract. So the current global price weakness is really about exports in the second half of Q2 19, rather than affecting today’s boats on the water, unless those cargoes are from Cheniere’s own capacity at its facilities. Still, the softness in pricing, while it has left the arb open, has possibly encouraged Cheniere to undertake its train maintenance fairly early in the shoulder season. Since March 22, maintenance at Sabine Pass T1 and T2 has limited exports to just three cargoes, which compares with an average of 6.5 cargoes leaving the facility weekly since the start of December 2018 (not including the fog events in February). The maintenance has been well-timed in terms of cutting cargoes destined to make a landing in the global shoulder season.
Even if global gas supply stays robust, Europe has a number of moving pieces in its market that could allow it to take the LNG that is expected to head its way (see North America monthly - Global LNG trends: Europe, the LNG sponge, 5 March 2019). First, European storage by end-October could be likened to the US at 4.0 tcf. End-October European inventories are likely to be at least 350 bcf (10 bcm) higher y/y. Europe also has the potential for fuel switching. It has a maximum of 459 bcf (13 bcm) of potential added gas demand from pushing hard coal out of merit this summer. However, achieving all of that substitution will be very difficult due to the need to manage peak demand as well as transmission constraints. Consequently, 318-353 bcf (9-10 bcm) of switching is a more realistic target. However, gas prices would need to fall in comparison to coal and carbon prices. Looking at the key pricing points at the TTF, the 5% fuel switching trigger (45% efficiency gas-fired plant vs 40% efficiency coal-fired plant) is now around 13.6 €/MWh ($4.50/mmbtu) after last week’s moves in European coal and carbon prices, while the parity trigger is now down at 11.6 €/MWh ($3.90/mmbtu). At current pricing, the 5% trigger is already being economically induced, and most switching should occur at that trigger. Given most of the fuel switch potential is in European markets where prices trade at a premium to the TTF, the TTF will possibly have to move lower than fuel-switch trigger levels to stimulate enough switching in the continental European market.
The next issue is how strong pipeline inflows into the EU continental gas markets will be. As gas prices start to drop towards the 5% fuel switch trigger, we expect pipeline flows will trend lower. Currently, we expect flows from the three NW European supply countries (the UK, the Netherlands and Norway) to drop by a combined 230 bcf (6.5 bcm) y/y over the coming six months. Depending on exactly how much LNG does get exported to Europe, that means flows from Algeria and Russia will need to drop by 280-350 bcf (8-10 bcm) y/y over the summer if the EU market is to balance and take all of the available LNG. We think Algerian flows are likely to fall by around 176 bcf (5 bcm) y/y and Russian supply could shed some 106-180 bcf (3-5 bcm) y/y. That Russian number is fairly low—only a 4-5% y/y drop—as we think that Russia will actively try to sustain its market share in the face of much lower customer demand and the competition from LNG.
The Henry Hub summer 2019 strip is currently around $2.80/mmbtu, while freight rates have consistently fallen and are now around $30,000 /d. Given those current inputs, we estimate the arb will close when European prices dip below $4.15/mmbtu (12.5 €/MWh), which is the delivered price into NW Europe (TTF) at the Henry Hub price, increased by 15% for liquefaction costs (to $3.20/mmbtu) plus transport costs of $0.60/mmbtu and regas costs of up to $0.30/mmbtu. Of course, what is variable or fixed cost for a particular market participant is also key. For those with vessels under long-term charter, which is the case for much LNG trade, that cost is not variable and requires a narrower spread. The variable vs fixed factor applies to regasification capacity as well. In other words, while we can say the arb closes at $4.15/mmbtu, that is not a uniform price for all.
The EU fuel switch trigger levels and the LNG arb closing level point to a convergence of equilibrating prices around $4.00/mmbtu, where demand-side increases in power are exhausted and LNG supply into Europe would start to decline. Of course, leaving cargoes in the US is likely to have a bearish impact on Henry Hub prices, although how much of an impact depends on the length of time that arb window is closed. But it does raise the prospect of a global race to the bottom, with the TTF following Henry Hub down while the JKM is pressed into following the TTF.
Currently, our reference case assumes no US shut-ins this injection season, power burn up by 0.5 bcf/d y/y and Henry Hub prices at $2.80/mmbtu. If supply is shut-in in the US, we think some will be directed into storage and some burned off in the power sector. If prices were to fall to $2.70/mmbtu, we estimate another incremental 0.7 bcf/d in power burn versus our reference case. On a monthly basis, that would essentially offset one shut-in train, which in turn would take about five cargoes (0.3 Mt) out of the market each month. A price decline to $2.60/mmbtu would offset 1.5 bcf/d of incremental gas demand versus our reference case, effectively taking about 10 cargoes out of the market each month (0.6 Mt). At most, we expect 96-144 bcf (2-3 Mt) of LNG could be at risk of being shut-in over summer 2019. That range suggests a maximum loss of two trains during the months most oversupplied. If we assume that loss is sustained over a three-month period, that works out to be 1.0-1.5 bcf/d less LNG feedgas demand that otherwise expected. That level is likely to be absorbed in the power sector, with some volumes flowing into storage too.
Obvious downside risk to pricing for the summer exists, even under normal weather. A milder-than-normal summer could exacerbate the downside risk further if the European market does indeed look like it cannot absorb all the volumes headed to it.
|Fig 1: LNG breakevens to NE Asia, $/mmbtu||Fig 2: Projected EIA weekly storage change, bcf|
|Note: Prices are based on futures on 22 March.
Source: Refinitiv, Energy Aspects
|Source: EIA, Energy Aspects|