The negativity at the end of 2018 was palpable for all risk assets, not just for crude. But we are still struck by the indifference of traders and the lack of uplift in flat price relative to the material tightening that is taking place in the physical oil market right now, which we noted in Perspectives: Heavy heart, 21 January 2019. This divergence has only accelerated in the last two weeks. US sour crudes are now among the most expensive grades in North America (only LLS is stronger), the contango in cash WTI has nearly disappeared, and all USGC grades are backwardated. This historically has not been bearish.
But at the same time, the signals from the product markets are confounding. Gasoline continues to deteriorate materially as the cracks compress and the spread between RBOB and distillate is at a seasonal low. The outlook going forward does not inspire much confidence, with high stocks and more supply expected in the coming weeks. Given the need to make more distillate in a world constrained by IMO spec changes, together with a fairly gloomy outlook for gasoline (see Light ends Outlook: Down by the bay, January 2019), these are at best conflicting signals for the market.
These trends are the result of a structural problem in the current oil market: too much light sweet crude production and too few sour barrels. Given the deterioration in gasoline margins in 2018, one would have expected to see a material shift in refinery yields away from gasoline, but the yield shift thus far has been modest and underwhelming. This is more likely to have been by default than by design due to price signals.
The strengthening in the sour crude market is sending just the wrong signal to refiners at the wrong time, as continued strength in sour markets could eventually lead to run cuts at complex refiners at a time when the market clearly needs to make more diesel in anticipation of IMO. It is not yet clear to us how the market can square this circle, but the more gasoline margins compress, the more distillate margins need to rally to carry the overall margin.
In 2018, oil price volatility was very much a function of quantity being affected by Iran sanctions, OPEC curbs and US production growth. But in 2019, while there are still plenty of potential pitfalls for supply, quality issues will be as important as quantity.
|Fig 1: Mars-Maya Coking margins, $/b
||Fig 2: LLS-Mars spread, $/b|
|Source: Argus Media Group, Energy Aspects||Source: Argus Media Group, Energy Aspects|
Sour crude markets are on fire right now. We have long been bullish on sour crude balances, but the extent of the tightness gripping sour markets today was only expected in Q2 19, by which time we had expected sweet-sour differentials to flip and sour crudes to trade consistently above their sweet counterparts. Yet Urals has already been trading at a premium to Dated Brent for a few weeks (before dropping slightly below yesterday) despite widespread sweet-sour switching among European refiners. But nowhere is the tightness in mediums/heavies and sours more visible than the US Gulf Coast. Heavy Louisiana Sweet (HLS) has flipped to a premium to Light Louisiana Sweet (LLS), which in turn is barely maintaining a $1 premium to Mars, while Mars is trading at premiums to MEH. In Asia, Oman’s premium to Dubai has been largely steady, while SOMO managed to sell both Basrah Light and Basrah Heavy at a premium of $1.50 to OSPs in its latest tender. All sour crudes are now backwardated and the increasing mismatch between sweet and sours that we have been highlighting for well over a year now is becoming extremely visible. The only problem is that this strength has come much earlier than we anticipated. This is partly down to the Canadian production cuts closing rail and pipe arbs of heavy crudes to the US, Saudi and Iraqi loadings to the US also being set to plummet to a record low in January, and now the potential threat of sanctions on Venezuela.
Structural declines in Venezuelan, Mexican and other sour crude producers and sanctions on Iran have been tightening sour crude balances since 2018. This year, the tightness has been exacerbated, particularly on the US Gulf Coast, due to OPEC cuts focussing its export cuts on the west and production cuts mandated by Alberta to ease inventory pressures. Indeed, the Canadian production cuts (see E-mail alert: Canadian production losses higher than expected, but backwardation to trump closed spot arbs, 12 December 2018) have closed rail and pipe arbs of heavy crudes to the US. Meanwhile, Saudi and Iraqi sours loaded for the US hit a 47-month low of just 0.86 mb/d in December, according to Kpler cargo tracking, while January-to-date loadings are a mere 0.43 mb/d and we estimate full-month loadings could be below 0.4 mb/d. Loading programmes for February suggest the lower trend will continue.
While there is a possibility that Canadian cuts may be reversed by April following the price rally, if excess inventories have been eroded by then, sour crude markets will remain tight as the nearly 2 mb/d of new refineries starting and/or ramping up this year are all effectively designed to run on sour crudes. By mid/late Q2 19—as Rongsheng, Hengli and Rapid (combined capacity of 1.1 mb/d) ramp up—medium sours could consistently trade above light crudes and Dated-Dubai is likely to flip, especially as refinery maintenance, which is biased towards sour crude refineries globally, will have ended by then. At the same time, should the US administration force buyers to cut Iranian exports by a further 50% from early May to around 0.7 mb/d, sour crudes in the east will tighten further. And with Saudi Arabia’s sour crudes tied up in new refining systems, buyers will struggle to find all the alternative barrels from the Kingdom alone, even if OPEC+ cuts are reversed. This is because new refining demand for Saudi crude amounts to at least 0.6 mb/d, while the average export reduction from Q4 18 levels of 7.6 mb/d is around 0.5 mb/d. In November 2018, Aramco did push to nearly 8.3 mb/d of exports but that was by surging production and destocking and hence we do not believe sustainable for an extended period. In fact, sours may tighten to the extent that sour crude run cuts may be needed to balance that part of the market.
At the same time, our calculations show that the sour-to-sweet switching ability of USGC plants has largely been exhausted. Indeed, in the USGC, rising tight oil supply over the last three years has priced itself into the local refinery system and led to a nearly uninterrupted downtrend in most local sweet-sour spreads, including LLS-Mars and MEH-SGC. Many refineries continue to adjust their operations to process more lights given the growing supplies locally (and with each turnaround cycle, switching capabilities will rise), but not nearly enough to absorb all the incremental production being added, and thus the region is still overwhelmed by additional light supplies.
Average crude input qualities reported by the DoE for the USGC show a recent jump to over 33 API (latest data for October 2018) from below 32 API just two years ago, and average sulphur levels have dropped from over 1.5% to below 1.4%. While refinery works undoubtedly play a role in these trends, inferred data suggest more light sweet crude is already entering USGC stills. Sophisticated refineries will always be able to adjust slates given enough time, but for now it appears difficult for sweet-sour spreads to widen materially until OPEC sour supplies return. In other words, even if OPEC opts to extend the production agreement in April, it will still have to flip around the quality of the cuts entirely and focus on raising sour crude exports to ease these quality tensions. This is particularly true given renewed focus on potential sanctions on all US imports of Venezuelan crude imports. This risks eliminating around 0.4–0.5 mb/d of heavy crudes entering the USGC (see E-mail alert: Risk of US sanctions on Venezuela's exports is rising; potentially bullish for sour crudes, freight and fuel oil, 24 January 2019).
So if refiners simply cannot get hold of sour barrels and cannot switch to sweets, run cuts will be necessary. The problem is that this is going to further complicate matters for products markets as margins for many complex refineries have been crushed and if they start cutting runs, it will only exacerbate the tightness in diesel markets. We have argued for some time that the world is facing a growing imbalance in crude quality as medium sour crude production outside of North America is largely in decline and is being replaced by light sweet US shale crude. As a result, two questions are on the lips of many: does the slate of crudes available to refiners lead to sub-optimal product yields given consumption trends leading to too much gasoline and naphtha instead of diesel; and more importantly, are refinery yields being forced towards lights as a consequence? The evidence so far is compelling, but it is not conclusive, and it hardly refutes the notion that crude quality is a major concern for the products market. The case could be made that the products markets is actually starting to signal to the crude market that there is less need for light sweet crude, and more need for medium and heavy sour barrels.
Brent: Prompt Brent spreads had been rallying steadily from very low levels since the start of the year, but that rally slowed this week, perhaps related to positioning in the WTI-Brent spread (WTI spreads have unexpectedly firmed this week). But given that prices were being aided by temporary factors such as weather disruptions and production outages, it was hard to get overly excited by the move in spreads, especially with refinery turnarounds just around the corner. The tightening in the sour market is clearly ringing alarm bells, however, and perhaps is indicative of fears in the market that more sanctions against Venezuela are coming. Cracking margins across the Atlantic basin continue to deteriorate, though we are not yet at run-cut levels. As mentioned above, the strength in the sour market could risk run cuts going forward at a time when refiners should be making as much diesel as they possibly can in preparation for IMO. In either case, there looks to be an oversupply of gasoline relative to distillate.
|Fig 3: WTI Dec-19 vs Dec-20, $/b||Fig 4: Brent and Dubai prompt spread, $/b|
|Source: Bloomberg, Energy Aspects||Source: Argus Media Group, Energy Aspects|
WTI: Even though US crude stockbuilds are continuing, the focus has shifted to prompt shale production, as signs begin to emerge of the Q1 19 deceleration predicted by oil service companies since late last year. Last week, total US rigs dropped, frac spread counts have fallen, proppant loading rates look light in the Permian, and Capex among Permian producers has been reduced since the Q4 18 sell-off. Meanwhile, Midland differentials have rallied, suggesting, in part, that Permian production growth is now slowing. In turn, this has supported WTI timespreads and WTI-Brent as less Permian crude is expected to move towards Cushing. We still project US crude production will grow y/y by 1.2 mb/d on average in 2019, but a big chunk of that growth will come in H2 19, after key pipelines and docks are commissioned. The importance of these pipelines (and also new export facilities on the Texas coast) to global balances is not at issue, but the timing of their commissioning is increasingly uncertain and will be meaningful for spreads and differentials.
If pipelines can be brought on sooner than expected it would remove one source of supply into Cushing, as more Permian crude would move directly to the USGC rather than to Cushing. Given the lack of Cushing builds in January and the way the physical market is currently trading,it appears the market is fading the risks of tank tops at Cushing in Q3 19. However, the timing of these pipelines remains uncertain and will be very meaningful for spreads. (for more details see our Data review: Department of Energy, 24 January 2019).
Products: Though the lightening of the crude slate and the imbalance between gasoline and distillate has been underway for a number of years, it is only now beginning to overwhelm the market. In 2016, 2017 and even 2018, distillate stocks were drawing counter-seasonally despite record refinery runs. But, over that time, especially in 2016 and 2017, refiners still had strong incentive to make gasoline as demand was very robust. Going forward, we will find out how much flexibility refiners around the world have to shift their yields further. Given IMO, gasoline balances, and stresses in the refining system, there is every incentive in the world to maximise distillate yields while minimising gasoline yields. In the US, distillate yields had been rising over the last 40 years as more and more heavy sour barrels were introduced into the US refining system. This stopped in 2010/11, roughly about when US light tight oil product began to surge. Since then, distillate yields have failed to increase materially and have for all intents and purposes flat-lined.
In Europe, we have seen a similar process. In 2017, we started to see distillate yields also run into resistance, even though the market was still telling refiners to make more distillate. This year is likely to give us a little more clarity on how refiners can manage this process. There is an unambiguous signal at the moment for refiners to shift yields, much like last year. If they don’t, then this is not just a signal from the product markets to stop making gasoline, but to stop producing light sweet crude.
|Fig 5: Dec RBOB vs HO, c/g||Fig 6: RBOB vs Brent crack spread, $/b|
|Source: Bloomberg, Energy Aspects||Source: Bloomberg, Energy Aspects|
Currently, the spread between RBOB and heating oil are at seasonal extremes. Given the tension in the sour crude market, there is a risk of run cuts. If the latter occurs, this will only serve to further tighten the distillate market, but it is not clear to what extent gasoline will clean up. In fact, at the margin, gasoline is likely to get worse as refiners will swing partially or as much as possible to lighter slates due to the dearth of sour crudes. Summer grade is a different beast, but with plenty of material allegedly stored from now, it is still difficult to get excited about these cracks despite them languishing at near record low levels (only 2011 and 2008 were weaker in the last decade). Winter grade is even worse, as the availability of incremental products that can find their way into the gasoline pool increases. If diesel markets tighten materially in the coming months, giving refiners incentive to make more diesel, they will also make more gasoline. We feel refiners have two choices: they will either continue to run the crude available, exacerbating the oversupply in gasoline, even if they meet distillate demand (although even that is in question given IMO needs); or they will cut FCC runs, which in turn tends to lower CDU throughput. That may ease the oversupply in gasoline, but it will most assuredly overtighten distillates. In both scenarios, this should continue to pressure RBOB versus heating oil, especially looking into Q4 19, when the IMO 2020 impacts will be most acute.
For us, there is one very important question for the oil market going forward: can the world absorb incrementally more light sweet crude? The answer will be extremely meaningful—there is as not yet enough decisive evidence to prove that refiners cannot shift their yields, but we haven’t as yet seen enough evidence to suggest that they can.
Energy Market Strategist
Chief Oil Analyst