Storing Oil at a Loss? Analysis Akin to the Drunk Looking for His Keys Under the Streetlight
This article claims that “Oil traders are borrowing from banks to store crude at a loss.” Almost certainly not.
The calculation is based on contangos in Brent crude. But the vast bulk of the oil being stored at sea is in Singapore, and is not BFOE, from what I can gather. This makes all the difference.
This is an example of what has long been a puzzle in economics in which futures prices in a central market (like Brent futures) are at less than full carry but inventories are held at other locations, or in non-deliverable grades (in the central market). People are puzzled because they don’t take into account transformation costs. It would be impossible for some locations/grades to be at a carry that incentivizes holding inventory while other locations/grades are in backwardation if there are no frictions (costs) in transforming a commodity from one location to another, or from one grade to another: take the stuff out from storage in the market that is at full carry, move it to the central market that is in backwardation, and capture that backwardation (i.e., the premium for immediate consumption/delivery). But if there are frictions, the costs of transforming (e.g., shipping) the commodity to the central market may exceed the backwardation/premium for immediate delivery, making it unprofitable to make that transformation to capture that premium.
This means that a commodity that is costly to transport, or where there are bottlenecks in the logistical or refining processes, will have different forward curves at different locations (or for different grades). Where inventories are held, the price structure will cover storage costs: where the price structure doesn’t cover storage costs, inventories will not be held. Similarly, if some grades of a commodity are stored, the forward curve specific to that grade will cover storage costs.
A good example is Iranian heavy crude in the spring and summer of 2008. At a time when light sweet crudes (WTI and Brent) were in backwardation, and oil prices were reaching record highs, 14 or so supertankers of Iranian crude were swinging at anchor. Why? The demand was for light sweet crude to make low sulphur diesel to meet new European regulations: due to bottlenecks in refining in Europe (European refineries being unable to economically process heavy sour Iranian crude into LSD), there was strong demand for sweet crude that was cheap to refine into LSD, and little was in storage. As a result, WTI and Brent prices were high, and their futures curves were in backwardation. However, there was weaker demand for heavy sour crude because of its unsuitability for producing LSD. Also, many refineries that were optimized to process sour Iranian crude were down for maintenance. As a result, Iranian crude sold at a very large discount to WTI and Brent, and the forward price structure for that crude made it economical to store it.
Moving forward to the present, storing non-Brent crude in Asia can be economical even if time spreads do not cover the cost of storage Brent in NW Europe. Even though the lack of carry in the Brent market indicates a high demand for BFOE, the cost of transporting crude from Asia may make it uneconomical to ship it to Europe to meet the high demand for oil there. Moreover, the grades of crude being stored on tankers in Asia may not be competitive with Brent. Similarly, demand is not high enough in Asia to make it profitable to refine all of the supply there. So it is uneconomic to move it to Europe, and it is uneconomic to refine all of it. Therefore, store some of it in Asia even when Brent time spreads are narrow.
Furthermore, the apparent squeezes in Brent mean that some of the demand for it is artificial, and that Brent spreads do not reflect the competitive economics of storage. That is, if there are squeezes or even anticipations of squeezes, Brent calendar spreads are artificially high due to the exercise of market power. It is particularly misguided to use Brent spreads to evaluate the economics of non-Brent storage in this case. (The major reason that squeezes can work is that it is impossible to transform non-Brent crude into Brent.)
Oil traders operate on very thin margins. They are not going to make uneconomic trades. Period. If they are storing oil on ships in Singapore, it is because it pays to do so.
So why the claim that they are storing at a loss?
It’s like the old story of the drunk looking for his car keys under the streetlamp, because the light is better there. It is easy for outsiders to observe Brent spreads, or WTI spreads: just look at a Bloomberg screen, or even futures settlement prices. Singapore spreads, not so much. Traders search to collect information about prices and price structures in an opaque market like Singapore, and can use that to evaluate the economics of storage opportunities. But you or I or journalists or even analysts at Morgan Stanley have a far tougher time doing that.
So the outsiders look under the Brent futures streetlight, and conclude that storing oil doesn’t pay. But the oil isn’t under the streetlight. It’s being stored in the dark. And those who can see in the dark are storing it there because they can see that it pays.