Out there in the cloud, somewhere in the e-bowels of an upstanding company’s financial reporting cesspool, is a curious commentary about future decommissioning liabilities. Under a “Provisions” subsection, this pipeline company’s annual report discloses that a provision for future liabilities must be recognized when, among other things, “the obligation can be reliably estimated.” That is the standard as set down by auditors, and the section drones about the process until the punch line where it is noted that decommissioning obligations are being “accreted over a period of 40 to 290 years.” What the…
There is no point in naming the company; the intent isn’t to embarrass the poor accounting department for adhering to the obviously questionable sanity of generally accepted accounting principles. Other pipeline companies likely have similar disclosures in their numeric tailings ponds, and I’d reference them also if I bothered to find them, but you try reading through all that crap. It’s like looking for your car keys in the muck after Coachella. And frankly, the odds are quite high that, after the insiders and auditors proudly signed off on the non-fiction tour de force, the section in question will never be read by a human again.
The interesting point though is the satisfied clicking-of-the-mental-machinery that would lead anyone to conclude that a 290-year projected decommissioning date has been “reliably estimated.” An estimated decommissioning date of “more than 50 years” is reliably estimated. One that implies the precision of a specific 10 year interval is a success to a bureaucrat and a joke to everyone else. But that is the way most things work now; we have our models that we simply plug numbers into, and then marvel at the outcome.
Sometimes this is inconsequential. We know an oil sands pipeline will last a very long time, but have no possible clue as to whether that will be 150 years from now or 300, or if the hated beast even makes it into triple digits. But the hard and fast model required inputs, so it got some, and it spit out a number with an implied precision that no one takes seriously. “Upwards of 100 years” would have gotten the point across, but that can’t be plugged into a model. An acquaintance helped decipher the calculation for me; apparently it is the mid point between high and low estimates of when the oil sands’ known reserves will be drained at present production rates. Or something like that. It doesn’t really matter.
Unfortunately though this phenomenon happens on a larger scale as well, where it really does matter. The most obvious is climate change. We assume that the sheer calculating power of the hardware together with the genius of the models (and their creators) can predict whether temperatures will rise by 3 degrees or 4, fifty years out. The math titans that started Long Term Capital Management in the 1990s believed the same thing, that their models had the financial world figured out, and they nearly blew up the global economy when their hedge fund imploded.
A decade ago, the financial world nearly collapsed again; there were many causes for the disaster but one large one was the US Federal Reserve’s view that low interest rates were nothing but a boon for everyone. All these models presumed to have the world figured out; the financial ones were spectacularly unsuccessful and there should be at least some second thoughts about the climate ones.
It happens on intermediate planes as well. People take a few known variables, such as installation of solar/wind facilities, sales of electric vehicles or installations of charging stations, and making headline-grabbing announcements such as that there will be no internal combustion engines by 2030. One guru of this line of thinking is Tony Seba, who is continuously referred to me. He’s founded something called ReThinkX, and publishes “compelling, impartial, data-driven analyses” that are as compelling, impartial, data-driven, and as stupid as the 290 year decommissioning estimate.
The pipeline accountants came up with their three-century estimate, to the nearest decade, because they had to. Accounting guidelines say that an estimate must be made, and one tied to the projected useful life of the oil sands is going to be a bit of a long shot to put it mildly. But the model spit out a number that served its purpose. And, fundamentally, we can rightly say who cares. There will never be a material impact on anything from that number. But the likes of Seba use their similarly robust models to capture and twist headlines and opinions.
The pace of modern media enables this loop. We are in a state where we have models to predict everything. All we need are the proper inputs, or so we think. We ask ourselves a simple question, or to use the above example, the gods of financial discipline ask of a utility, what is the expected useful life of your pipelines? You raised capital based on some perceived demand and built the things for a purpose, so simply tell us how long a period to amortize them over and we’ll give the thumbs up to your financial statements and all will be well.
Sometimes these numbers create what behavioral economists call “anchor points” which get lodged in the psyche through continual reinforcement by the media (admittedly, obscure page-57 estimates of decommissioning liability dates don’t get all that much press). We hear all about Seba’s theories of the imminent death of fossil fuels, and because he’s a Stanford University professor his babble gains instant credibility. If his models spit out sensation results, it gets headlines and shapes opinions. By contrast, the pipeline company’s estimates, if you ever encountered them at all, would not give you cause for alarm, and you would simply think what a weirdly specific estimate that is obviously unknowable.
Unfortunately, Seba and his anti-petroleum ilk come across as valid prophets due to their pedigree, the fear they engender, and their tactical use of media. The fact that they viciously abuse the credibility gleaned from their academic posts (Seba’s presentation outlines his use of the “Seba Technology Disruption FrameworkTM” – yes, it is trademarked) to advance their own glory-hounding is not noted anywhere; that sort of talk could only come from a denier (to borrow their vulgar, disrespectful and gleefully stupid term).
The world is changing as green energy becomes more prevalent. But the ripples and impacts of switching the planet’s primary energy sources are not understood at all. No one’s model can predict that. When you come across pompous prophets hiding behind impressive credentials and gunning for glory, remember that even the unconditionally unglamorous world of accounting creates impossible predictions. So those that come from self-promoters should be treated with the same consideration as Russian-dating-site spam.
Read more insightful analysis from Terry Etam here