A structural shift in the WTI/Brent premium
The WTI is usually called light, sweet oil given its lower content of sulfur and low viscosity. The easier cracking process associated with those properties has historically turned into a structural premium vs. Brent prices of 1.2 USD since 1985 (first chart below). The spread has shown some volatility over time, but has always tended to revert to the mean. Since the outbreak of the crisis, the spread has shown a long lasting period of discount that has to be explained.
Going deeper into the idiosyncrasies of both markets, one should bear in mind that the WTI is more local than the Brent. Brent is deemed to be much more influenced by global factors, such as worldwide trade or Asian demand. Even though USD trends should affect Brent, and as many producing-countries use it as benchmark (Russia and Nigeria, for instance), the oil/USD relationship remains an issue whatever the oil marker retained. It is supposedly negative because lower USD entails stronger demand from non-USD users, but the correlation shows variability over time.
Looking for the usual suspect
We list below the potential drivers of the WTI/Brent spread. Beforehand, we show that as with many spreads, the WTI/Brent differential shows a clear cut directionality. Historically, rising WTI prices would go hand in hand with a wider spread. The recent correlation break is particularly striking.
Global Trade. Even though the global export/Baltic Dry Index relationship has weakened somewhat over the last few quarters, there is a striking link between the WTI/Brent spread and the BDY as the chart above shows. Yet, the correlation may be spurious as a weaker trend in global trade should affect the Brent primarily.
Inventories The WTI is quoted for delivery at Cushing with inventories being stored onshore in Oklahoma. Brent depends much more on cargo traffic and availability in the North Sea. The chart above shows a weak but negative relationship between the WTI/Brent spread and the inventories at Cushing. The second chart also highlights the strong and positive relationship between the WTI contango (12-months contract vs. first nearby contract) and inventories (their levels in relation to 4-year average). This well-known relationship between inventories and the slope of the futures curve is once again verified.
Yet, tracking the maritime flows of cargo in Europe is impossible. We could therefore try to explain the current WTI/Brent spread by a differential in contango/backwardations in both markets.
Contango/backwardation is a reflection of the relative scarcity of commodity products, but is also the main driver of returns. Commodity returns depend on the changes in spot prices and on the roll/riding of the futures curve. The steeper the contango, the more negative the roll return.
As can be seen on the chart below, whenever the WTI contango is steeper than that of the Brent, the WTI/Brent spread tends to shrink or become negative. The roll return for WTI is currently highly negative in comparison with that calculated on the Brent futures curve. The steeper contango of the WTI is therefore offset by the discount against the Brent.
A comparison with the FX market may be useful. According to the covered interest rates parity, yield spreads explain spot/forward differences. Here, the ex ante roll return (one of the components of the excess return - sum of spot and roll return) translates into a future price differential.
Refinery margins Refining margins are proxied by the 321 crack spread which measures the income generated by the refining product minus the cost of the oil input. 3 barrels of crude are supposed to generate 2 barrels of gasoline and 1 barrel of heating oil. As can be seen above, the spread has reached a high level – suggesting that refinery margins are still very high in the US. This should entice refineries to increase their production of gasoline or heating oil.
Yet, the relatively low level of capacity utilization, combined with excess supply of distillates and gasoline, suggest that the WTI could remain low. Especially since the huge contango may entice producers to stockpile and sell forwards.
The widening of the spread is thus mainly a US based story (even though the spread may also have widened by the unwinding of mean reversion driven long WTI/short Brent positions over the last few days - see chart below).
The widening of the WTI/Brent is widely explained by US idiosyncratic factors. Contrary to the 2009 episodes, when wide refining margins enticed oil companies to produce more distillates, the situation is clearly different today as gasoline inventories are very high. (Technical factors such as the high level of shipping costs within the US (East Coast, for instance) are also suggested. It adds to our view that as long as there is no exit for excess gasoline inventories by sea or pipeline, the WTI glut will remain).
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.