The first half of 2009 saw weak demand relative to supply, and the front month was trading on average at a discount of -4.58% to the second month on the curve.
Due to the economy (etc. etc.), large crude oil players such as Glencore, Morgan Stanley, Vitol et al. were paid a premium to store crude oil for the future, when demand would recover. And store they did — floating storage rose to 100 million barrels and transfers from PADD 3 to PADD 2 averaged 36.7 MMbbls/month.
When the economy began to rebound at the start of this year — the opposite held true. January 2010 saw a significant tightening of the contango with the front month trading at just a -0.74% discount. Floating storage decreased to around 20/30 million (exact figures are difficult to obtain given the proprietary nature of leasing) and transfers from PADD 3 to PADD 2 dropped to 31.8 MMbbls, 19% lower year-on year.
- Energy Futures Now: Crude, RBOB, etc.
But how precise is the relationship between the contango and the change in transfers as demonstrated in the Chart of the Day in today’s issue of The Schork Report? And does a large increase in transfers beget a steeper contango or vice versa? Intuitively, we assume that macro-economic factors will signal whether demand is strong or weak relative to supply, and speculators will use this to bid up/down the front month. Producers will then adjust their storage levels accordingly.
For 2009 we saw transfers from PADD 3 to PADD 2 peak in July, at 41.4 MMbbls and then fall to 32.7 MMbbls in August, while the front month was trading at a discount of -1.49% in July and -2.19% in August before declining in September. This implies that the relationship is actually opposite to our assumption and lagged by ~ one month. Traders need to see a drop in storage before they are willing to start bidding up the front month — i.e., the fundamentals are driving the speculators instead of the other way around.
The one month lag is borne out by correlation. The correlation between changes in crude oil transferred and the front month discount stands at 0.30 when no lag occurs, but rises to 0.35 when a one month lag is introduced.
Recently, March and April have seen huge builds in the Cushing, Okla. hub. In fact, according to the DOE’s weekly petroleum reports, April 2010 saw the largest cumulative build at the Cushing Hub ever recorded for the month. The key take away here is that the steep contango is poised to hold well in to May. The Deepwater rig spill is not, repeat is not, the game changer needed to push the front month higher. Analysts at The Schork Reportare looking to see either a steady decline in builds or some other market moving news before we look for the futures curve to become flatter.
Stephen Schork is the Editor of The Schork Reportand has more than 17 years experience in physical commodity and derivatives trading, risk systems modeling and structured commodity finance.