I tend to avoid blogging about the stuff that I teach about in my class. That is probably a mistake.
A recent presentation by J. David Hughes about shale gas and tight oil has gotten some recent attention. It makes two conclusions. First, the price of shale gas is going to go up. Second, the tight oil boom is a chimera.
The conclusion that the price of shale gas is going to rise is true. It is also not news. (Why was it treated as such?) For example, I made a presentation on October 4th, 2012, to the Business and Environment Initiative at HBS on fracking and shale gas. In that presentation, I pointed out three things:
- If you assume land costs of $5000 per acre and opex of $1.50 per MMBTU, then the “point forward costs” (i.e., marginal costs including land acquisition) of shale gas are $8.31 in the Fayetteville shale and $8.75 in Haynesville;
- ARC Financial estimates that $22 billion per quarter in capex is needed to maintain production, up from current expenditures of $12bn;
- Shale gas drilling has gone off a cliff (rigs in operation have crashed from a little over 1500 at the beginning of 2008 to only 500) ... and given the well-known hyperbolic decline rates of shale wells that meant production would soon decline unless prices recovered. (I reference those hyperbolic declines in this January 2012 post on Argentina.)
In short, Hughes is correct (his back-of-the-envelop calculation of the needed capex is actually only half that calculated by ARC) but his conclusion is not news and should not be treated as such. There are industry professionals who claim that well costs will fall and that land costs are not sustainable, so the marginal cost data should not be taken as data from on high. (Chesapeake has said that its magic spot for the Marcellus shale is around $5.) That said, few believe that natural gas prices will remain below $4 per MMBTU for much longer ... but “much longer” probably means more than a year. I do not expect U.S. prices to suddenly jump.
What about tight oil? Well, we already know that the marginal cost of tight oil is at least $60. When you add in the realization spread (the difference between the companies’ realized revenue per barrel and some benchmark price, usually Brent) you get a marginal cost of at least $90.
Hughes goes on to argue that the Bakken will peak in 2017 and then decline, but that rests on some dicey assumptions. First, that the 2010 EIA estimate of 9,767 potential wells remaining is correct. It is probably not: it assumes spacing of 2 wells per square mile, when four is already occuring and eight is already feasible. Second, it includes only discovered reserves. Finally, it assumes that oil prices will not rise further. The first two are relatively improbable.
Moreover, the Bakken is only one of many plays in the United States. It will start to decline relatively soon, if not by 2017 ... but other plays are expected to increase. The EIA really does know what it is doing when it makes its projections.
In short, it is true that the Citigroup analysis seems breathlessly out to lunch and it is true that tight oil plays require high prices. But we are not facing peak-tight-oil anytime within the United States. And while natural gas prices are fated to rise, the rise is not immediate.