In an article from earlier this week, I described what contango is and what it means for oil-linked exchange-traded funds. The gist of the article was that these ETFs must constantly roll their oil futures positions from one month to the next as contracts expire.
When the futures curve is upward sloping―with prices higher the further out in time you go―it's called contango. When the futures curve is downward sloping―with prices lower the further out in time you go―it's called backwardation.
As I explained, when the futures curve is in contango, as it currently is, oil ETFs end up with fewer and fewer contracts as time goes on as they sell cheaper-priced contracts and buy more expensive contracts.
Over time, fewer contracts translates into losses for ETF investors unless front-month oil prices themselves continually march higher.
John Hyland, formerly the chief investment officer of United States Commodity Funds―the issuer of popular ETFs such as the United States Oil Fund (USO | B-100) and the United States Natural Gas Fund (UNG | C-94)―doesn't believe that this "volumetric" approach to analyzing contango is the easiest for investors to understand.
Rather, Hyland offers an alternative take on how best to describe the ETF contract rolling process, which you can read below.
Hyland: I have been trying to steer people away from describing backwardation and contango in "volumetric terms" for the last 10-plus years. By “volumetric,” I mean referencing the change in the number of contracts an investor or fund holds, as opposed to the financial impact.
I have done this because I have found that to RIAs [registered investment advisors] and other ETF users, volumetric descriptions don't really tell them anything useful. This is because a) ETF investors do not actually care how many contracts there are; and b) backwardation and contango do not actually tell you if you make or lose money, only how you do relative to spot price movements.
Instead I give them something that describes rolling contracts in backwardation as buying the new ones at a "discount to the spot price," while rolling them in a contango market involves buying the new contracts at a "premium to the spot price."
Since RIAs understand what happens to the premium or discount when you buy bonds―you write off the premium and accrue the discount―it is easier for them to understand what actually is happening here. An investor gains relative to the spot movement from buying at a discount and loses relative to the spot movement when paying a premium, just like with bonds.
Examples:
"When prices of further out oil contracts are lower than the current spot month price, this is called 'backwardation.' The financial impact of backwardation is that when an oil ETF rolls to the next month, it buys that contract at a discount to the current spot price. Over the remaining life of the next-month futures contract, the ETF will outperform the movement in the spot price by the amount of the discount, although there is no guarantee that overall they will make or lose money on this position."
"When prices of further-out oil contracts are higher than the current spot month price, this is called 'contango.' The financial impact of contango is that when an oil ETF rolls to the next month, it buys that contract at a premium to the current spot price. Over the remaining life of the next-month futures contract, the ETF will underperform the movement in the spot price by the amount of the premium, although there is no guarantee that overall they will make or lose money on this position."
My use of the phrase “premium and discount” runs counter to the use of those phrases by Professor Geert Rouwenhorst, who will describe the front month as trading at a premium if the market is in backwardation, or trading at a discount if the market is in contango―the exact opposite of my use.
I can see where Geert is coming from if you just think of it in terms of commodity economics, but his approach does not help an RIA whose primary knowledge of premium and discount is from bonds. Paying $41 a barrel for a WTI future when the front spot month is $40 will have the exact same effect as paying $101 for a bond, which will be worth $100 at maturity.
If you keep both investments to the end, you will write down the $1 premium. The reverse of course is true of buying at a discount. The only real difference between the futures and the bonds is, barring default, you know what the bond will be worth at maturity ($100), but you do not know what the spot month futures price will be at expiration.
Still, in both cases, what you do know is what happens to premiums and discounts.
It's also worth reiterating that the gain or loss from backwardation or contango doesn't necessarily happen the day the ETF rolls. Selling $100 million worth of contracts at $40, and buying $100 million worth of contracts at the same time at $41, does not create any gain or a loss at that time (if you ignore brokerage costs). After all, in both cases, you have something worth $100 million.
