After the announcement of the Seaway reversal back in November 2011, a development which some say was oddly anticipated by the market, the Brent-WTI spread collapse from a near record $30 to just $7 in the span of three months. Further alleviating tensions was the fact that Italy is now once again back firmly in control of Libyan Brent production. Yet recent developments in the Persian Gulf, and elsewhere, have led to the Brent-WTI spread trade becoming an energy trader's widowmaker yet again, as it has doubled from $10 to $20 as of early this morning in less than a month. What happens next, and what are the implications for the energy market as a result of the violent move wider? Here is Goldman's David Greely with some observations and some suggestions.
First, the take home:
The WTI-Brent spread has widened to -$19.02/bbl, its widest level since the announcement last November by Enterprise and Enbridge that they would reverse the Seaway pipeline to flow crude oil from Cushing to the US Gulf Coast (Exhibit 1). Following the November announcement, the WTI-Brent spread narrowed to -$7.93/bbl, even though at the time the reversal was not scheduled to be complete until April 2012 (and the completion date has since been pushed back to June 2012). We expected that this narrowing of the WTI-Brent spread would be supported by a build in crude oil inventories, with the build in inventories taking pressure off the Midwest-Midcontinent physical crude oil supply-demand balance and reducing the WTI discount. Although the anticipated build at Cushing has taken longer than we expected to begin, we are now starting to see crude oil rushing to Cushing from Canada and the Midwest. More specifically, crude oil flows on the Spearhead pipeline from the Midwest to Cushing have recently surged to near capacity levels, and crude inventories in Cushing are beginning to build, increasing by 1.5 million barrels last week according to the DOE.
We expect that crude oil inventories in Cushing will build rapidly in the coming months in anticipation of the Seaway reversal, with these higher inventory levels putting downward pressure on the WTI timespreads, like we observed in both 2010 and 2011 (Exhibit 2). Consequently, we are now introducing a trading recommendation to be short near-dated WTI timespreads, or more specifically, short May 2012–long June 2012 WTI futures contacts (at an initial value of $0.62/bbl.) Further, with the March 2012 WTI-Brent spread now trading at -$19.02/bbl, we are now closing and recommend taking profit on our short March 2012 WTI-Brent spread, long the December 2012 WTI-Brent spread trade recommendation (initial value $1.79/bbl, closing value $7.09/bbl, profit $5.30/bbl). We continue to expect the December 2012 WTI-Brent spread to narrow, but it may take the market some time to begin to re-price these longer-dated WTI-Brent spreads in line with pipeline tariff economics.
Keep an eye on Cushing inventories is the primary indicator of what happens next:
Midwest and Midcontinent prices spreads continue to direct crude oil to Cushing, and the flows have begun to respond in earnest In our October 2011 Energy Watch, we introduced a framework for understanding the price of WTI relative to Brent in term of the underlying regional price spreads that direct crude oil through the US Midwest and Midcontinent and on to the US Gulf Coast (see our GS Energy Watch: Between Scylla and Charybdis, October 4, 2011 for details). Specifically, we showed that in addition to decomposing the WTI-Brent spread into the LLS-Brent leg (representing the transatlantic light-sweet crude oil arb) and the WTI-LLS leg (representing the difference between Cushing and the US Gulf Coast), it is important decompose the WTI-LLS leg into a WTI-MSW (Canadian Mixed Sweet) leg and a LLS-MSW leg Following the Seaway announcement, we expected that WTI would price above MSW in order to direct the flow of crude oil to Cushing, where a contango in WTI futures prices would support an inventory build. By building inventories in Cushing, the pressure on the barge market to carry crude down the Mississippi from Wood River to the US Gulf Coast would ease, allowing the LLS-MSW spread to narrow. The changes in both the WTI-MSW and LLS-MSW spreads would contribute to the narrowing of the WTI-Brent spread. Once the Seaway reversal is complete, we would expect the inventory stored in Cushing to flow out of Cushing and down to the US Gulf Coast.
While directionally the narrowing of the WTI-Brent spread following the Seaway reversal announcement made fundamental sense, in our view the WTI-Brent spread narrowed too much in the front of the curve (for contract months before the reversal would be complete) and not enough in the back of the curve (for contract months after the reversal would be complete). The WTI-Brent spread narrowed too much following the Seaway announcement on November 16, 2011, in our view, because the LLS-MSW leg narrowed to the point where barge flows would slow to the point where too much crude would be remain trapped in the Midwest and Midcontinent, likely overflowing PADD 2 storage capacity as Canadian and Bakken crudes continued to flow into the region. Further, the WTI-MSW leg strengthened sharply, which we expected would lead to a large build in Cushing, which could potentially breach storage capacity prior to the reversal of the Seaway.
In the recent widening of the WTI-Brent spread, the LLS-MSW spread has begun to widen again in recent weeks to what we see as more appropriate levels, which will likely motivate more barge flows down the Mississippi. However, the WTI-MSW leg is also widening, and sending a stronger signal to send crude oil flows to Cushing. While the Spearhead flows were slow to respond to the WTI-MSW spread, recent evidence suggests they are surging and will likely drive a strong build in Cushing in coming months.
We find that net flows into Cushing via the Spearhead and the Ozark pipeline follow price differentials between MSW and WTI typically with a 2-month lag. Consequently, the strong price incentive that has been present since late November should lead to increased net oil flows into Cushing at least through February and March. Our model suggests that flows on the Spearhead pipeline should increase to close to its capacity of around 180 thousand b/d in February and March, up from just 80 thousand b/d in January. Pipeline flow data from Genscape reports that Spearhead flows have indeed picked up sharply since the beginning of the month, and are now close to 123 thousand b/d. Further, Bloomberg reports that shippers have requested space for 737 thousand b/d on the Spearhead in February, up from 93 thousand b/d in January, implying that the pipeline should continue to ramp up and run at maximum capacity going forward. Given the current price incentives, we expect that shipments on the Keystone pipeline from Steele City to Cushing will also increase, potentially at the expense of flows to Wood River and Patoka.
However, as flows on the Spearhead pipeline approach capacity, a new transportation bottleneck is emerging. Prices of Canadian crude and crude fed into the Canadian pipeline system have collapsed in recent days. More specifically, Western Canadian Select (WCS) is now trading roughly at a $35/bbl discount to WTI, compared to an average discount of $16.50/bbl in 2011. Syncrude, which traded at a $9/bbl premium over WTI in 2011 is now trading at an $18.75/bbl discount. The reason for the current dislocation can be found in the combination of growing supplies and unplanned refinery outages.
Canadian production ramped up rapidly in recent months, increasing exports of Canadian crude into the US Midwest. Weekly DOE import data shows that crude imports into PADD 2 and PADD 4 are up 350 thousand b/d over last year. At the same time, production from the Bakken shale in North Dakota has been growing at close to 150 thousand b/d year-overyear. While this rising production alone is likely not sufficient to create the bottlenecks that have recently resulted in the blow outs of the Canadian crude oil grades, the situation has likely been aggravated by a series of refinery problems in the Midwest. The most important of these is the unplanned maintenance work on a fluid catalytic cracking unit (FCC) at BP’s 405 thousand b/d Whiting refinery. The company announced on Monday, February 6, that the work has been completed and it is bringing the FCC back online. Between the rising production and refinery outages, there is a surplus of crude oil in Western Canada and the US Midwest. While Cushing inventories are still low and could accommodate this crude, the pipeline capacity to carry this oil from the Midwest into Cushing is rapidly approaching capacity limits, putting downside pressure on Canadian differentials. For example, the WTIMSW spread has exploded from under $10/bbl at the end of January to a record level near $25/bbl yesterday. The blow-out in the WTI-MSW spread coincides with reports that shippers have requested space for 737 thousand b/d on the Spearhead in February, far more than its 180 thousand b/d capacity, suggesting a severe dislocation. Consequently, there appear to be logistical bottlenecks not only to bringing crude oil out of Cushing, but to bringing crude oil into Cushing as well. While this pressure on Canadian grades should ease once the hiccups in the Midwest refinery system subside, ongoing crude production growth in the Bakken shale and Canada combined with limited pipeline capacity to ship crude from the Midwest to Cushing could mean that the reversal of the Seaway pipeline will remove one crude transportation bottleneck, but another may persist further north, something we will continue to monitor and analyze going forward.
In the meantime, the large premiums of WTI crude at Cushing over light-sweet grades traded in the Midwest and Canada will likely keep crude flows to Cushing at the highest possible level. On net, we expect crude oil inventories at Cushing to build strongly for at least the next two months and possibly until the end of May. We expect that Cushing inventories could build by 10 million barrels in the next two months, reaching 40 million barrels by the end of March, and potentially rising to 47 million barrels by the end of May, pushing storage capacity utilization rates well above 60% from currently around 45%. This will likely put substantial pressure on WTI timespreads over the coming months.
Finally, a look at refineries:
As a result, US East Coast refiners kept runs at exceptionally low levels and started to take down capacity more recently. Of the 1.6 million b/d of East Coast (PADD 1) refining capacity, 700 thousand b/d are currently in the process of being idled (Exhibit 8).
- ConocoPhillips began shutting down its 190 thousand b/d Trainer refinery in September.
- Sunoco began shutting down its 178 thousand b/d Marcus Hook refinery in December.
- Sunoco is also planning on idling its 340 thousand b/d Philadelphia refinery by July if a buyer cannot be found.
- Hess has recently announced that it is shutting the Hovensa refinery located in the US Virgin Islands, which has been a large supplier of gasoline to the US East Coast.
With Midwest refiners continuing to enjoy a strong crude oil cost advantage over their counterparts on the US East and Gulf Coasts, a surplus of refined products as well as crude oil is beginning to grow in the US Midwest. For example, motor gasoline in the Chicago area is now pricing more than $0.14/gallon below motor gasoline on the US Gulf Coast, flipping the normal pricing arb which would traditionally direct motor gasoline from Gulf Coast refiners to Midwest consumers. Prior to the recent refinery issues in the US Midwest, the discount of Chicago to US Gulf Coast motor gasoline widened to almost $0.30/gallon (Exhibit 9). This suggests that the US refined product markets are increasingly prone to the logistical issues and resulting dislocations that have affected the US crude oil markets over the past year. Further, with the ongoing closures of refineries along the US East Coast and limited logistical infrastructure to carry refined products from the Midwest and US Gulf Coast to the US East Coast, we expect that the motor gasoline markets, in particular, will be highly vulnerable to regional dislocations as we head toward the summer driving season.
Great. Now what happens to Brent-WTI if and when the geopolitical situation in Iran passes the boiling point?