Massive increase in the spread between crude oil and refined products

04 Sep 2017

The complexity of the problem of trying to understand what is happening in the oil market has increased dramatically this month with the advent of Hurricane Harvey.


This looks to be the most disruptive hurricane, so far as US energy markets are concerned, since hurricane Katrina in 2005. According to Bloomberg, it has so far led to around 4.5m b/d, or around a quarter of US refining capacity being shut down. The most immediate impact of this has been a massive increase in the spread between crude oil and refined products. This is because crude demand and refined product supply, both decline when refineries shut down. When something like this happens, it reminds us how complicated and interconnected the process of suppling the fuel we take for granted, really is. While the impact has been most strongly felt in the prices of US crude and gasoline, the effects have rippled around the world. The usual flow of shipments of US produced middle distillates into Europe has been interrupted and Asian crackers that process fuel oil into lighter products have stepped in to fill some of the gap leading to fuel demand which has supported short dated crack spreads.

The medium term impact could be quite complicated depending on how long refinery shutdowns last:

  • With refineries (as well as some ports and pipelines) closed on the US gulf coast, crude from shale producers that would normally flow to them has nowhere to go and at some point production has to be reduced.
  • The Hurricane will also tend to reduce economic activity and consumption of fuel in affected parts of the US.
  • The elevated price of fuel might subdue demand beyond the flooded affected areas and indeed, outside of the US.

So there are potential bullish and bearish medium term impacts for the overall supply and demand of oil. Even if refineries manage to restart quickly, the effects of the hurricane will add considerable noise to the weekly US figures which are the main set of reliable and regular information we have for supply, demand and inventories anywhere in the world (much of the other data relies on estimates and inferences). It is hard enough to separate core changes in the balance of supply and demand from seasonal ones, and that task has now been made considerably harder. The effects of the hurricane will first be seen in the figures to be published on 07 September, but could take a number of weeks to settle down.Elsewhere, the picture is also complicated. The market was caught off guard by the IEA’s monthly report published in August, which reduced their historic demand estimates for some developing countries, by a collective 400k b/d. While they kept growth forecasts for this year and next year broadly unchanged, the base they are growing from has been reduced. As a consequence, the report forecasted OPEC needing to produce 32.6m b/d and 33.0m b/d in Q3 and Q4 of this year, respectively, to balance the market. Unfortunately, that coincided with an increase in OPEC production which the IEA estimated to have risen by 230k b/d in July to 32.84m b/d. It saw compliance with cuts from the countries that signed up to them as having slipped and Libyan supply increasing, to the extent that if the July level for OPEC production continued for remainder of the year, it would lead to a market surplus of nearly 300k b/d in the current quarter and a small deficit in the final quarter. The figures painted a very different picture to the previous report which suggested significant deficits for the remainder of this year. However, early estimates of August OPEC production suggest that it has fallen back again as compliance has increased and Libyan production has been disrupted by unrest in the country.

Adding all this together, we get a very confused picture that will probably not become any clearer until we have seen another couple of months’ worth of data. In combination with the holiday season, it is perhaps unsurprising that there has been very little direction to oil markets during August. Brent came close to breaking 50 $/b per barrel, but could not quite manage it. Nor has it made convincing progress beyond 53 $/b. Over the last couple days however, the back end of the curve has showed signs of life. The December 2019 Brent contract has broken through its 200-day average and come close to 54.50 $/b. Hedgers who buy oil at the long end have been protected to some degree when the price of the front contract has risen, as the back-end of the oil curve has been kept in check by producer hedging. As consumer hedging activity increases in the autumn (hedging demand has been very thin over the summer) it could test the ability of producer selling to absorb long dated consumer hedging demand. Aside from the short-term uncertainties and vagaries of investor activity at the front of the curve, global growth seems to be picking up. The OECD recently increased is forecast for 2017 to 3.5% (the highest level in 6 years) and 3.6 % in 2018 and this would tend to underpin oil demand growth.