One week ago, we wrote that as a result of the collapsing crude contango, oil tankers (such as the fully loaded Distya Akula which has been on anchor in the Suez Canal for one month unable to find a buyer for its cargo so it continues to wait) "will soon have to unload their cargo", in the process flooding the already oversupplied market with millions of barrels of crude oil, thus pushing the price of oil far lower. But how many millions of barrels, and how much lower will the price of oil go?
For the answer we go to Deutsche Bank's Michael Huseh, who has done the calculations to get the answer.
What he finds is that since the start of 2014, global floating storage inventory has ranged between 80 and 180 million barrels (Figure 1). According to estimates of the global VLCC fleet at the end of 2014, the potential storage capacity is implied to be 1169 million barrels. Adding Suezmax vessels would add 528 million barrels of capacity.
After touching 186 million barrels in early March, inventories have begun to decline once more. Since the start of 2015, one can identify both periods when builds in floating storage have been associated with rising Brent prices, and also periods when draws in floating storage have been associated with falling Brent prices (Figure 2). Since the Arabian Gulf has represented much of the variability in floating storage inventory, one can also measure the incentives to add or withdraw from storage using Arabian Gulf tanker rates.
South East Asia would be another valid candidate to measure economics, as floating storage inventories in that region have moved in a very similar fashion (Figure 3).
As we discussed recently, as a result of a recent surge in hedging activity in the front-end of the strip, absent a dramatic collapse in spot prices, the contango is now so low as to make offshore storage no longer economical. Specifically, based on the all-in cost of operating tanker storage (dirty VLCC tanker day rates, financing, transit and transfer loss, insurance and bunkers, Figure 5), the current storage cost is too high relative to the steepness of the Brent forward curve. This means that prices do not justify inventory build, but rather gradual inventory drawdown as existing storage trades are unwound.
What is the current prevalent duration of booked offshore storage? A comparison of the historical profitability of storage trades of varying lengths indicates that even at the most extreme instances of contango in the last two years, the Brent forward curve is only steep enough over the first 2 to 6 months to justify the floating storage trade. Comparing the trade economics over a one-month horizon (Figure 4) and over a six-month horizon (Figure 6) shows the relative unattractiveness of the six-month trade. We use the second month Brent contract owing to discontinuities in the pricing of the rolling first month contract. Thus we would expect that floating storage trades begun in late January or early February would be unwound by July or August.
As DB calculated, comparing the current level of floating storage (157.3 million barrels) versus that in early February (126.6 million barrels), there may be an additional 31 million barrels of inventory to be drawn down between now and the next inventory trough over the next several months. Depending on the duration of drawdown (three months or six months) this could mean anywhere from 165-330 kb/d of incremental supply.
So how, according to DB, should one trade this imminent surge in incremental supply?
A tactical short position in Brent may benefit from the contango roll yield which over the first six months of the curve is an annualized 14%. Over the first year of the Brent curve, the roll yield is 11.9% p.a., and to provide an extreme comparison, the roll yield over the first six years of the curve is only 5.3% p.a. In other words, in a flat oil price scenario the contango roll yield for a short position would still provide positive returns if the curve structure remains static. In an upside oil-price scenario, the six-month forward contract should rise slower than the spot price.
Long WTI-Brent may be a viable alternative: because positioning in Brent is more clearly extended than NYMEX positioning in WTI, and also because US refineries returning from maintenance may add an incremental 717 kb/d of refinery crude demand between now and June, we believe WTI may be better supported than Brent. Brent net long non commercial positions rose to 164 thousand contracts in the week ending 15March, which is just below the 2015 high of 166 thousand contracts, although still some way below the 2014 high of 195 thousand contracts.
In WTI positioning on NYMEX however, net long non commercial positions stand at 331 thousand contracts, only 69% of the record high of 480 thousand contracts in June 2014. Therefore an alternative to selling Brent outright may be a long position on the WTI-Brent spread.
* * *
Of course, if DB's calculations are correct, and if over the next three months 20% of the total 157 million barrels in offshore inventory are set to come onshore, not only will underlying prices slide, but higher beta assets, such as energy equities but mostly junk bonds due to their record high correlation with energy prices as we showed before...
... the best trade may be to either sell cash bonds or, if one can find them in this illiquid market in which even the ECB is now actively involved in bond purchases, simply buy junk bond CDS.
Because between the surge in recent hedging, the collapsing contango, the failure of supply to decline, the failure of demand to increase, there is only one thing the price of crude oil can do: tumble, no matter how many flashing red "OPEC meeting" headlines Bloomberg blasts at idiot headline-scanning algos.