CryptocurrencyInvestingBanksReal Estate

A new report highlights the incredibly high environmental cost of Tesla’s Bitcoin investment

June 18, 2021, 3:00 PM UTC

In the 1942 film “The Man Who Came to Dinner,” legendary comedian Jimmy Durante tickled wartime moviegoers in a madcap number called “Did You Ever Have the Feeling That You Wanted to Stay.'” Jimmy’s seated at the piano, and the musical twist is that he doesn’t know if he’s coming or going. The bantam funnyman keeps bouncing to his feet and donning his stetson as if to leave, then sitting back down shedding the hat like he’s staying. While doing his yo-yo routine, he’s thumping the keys and trilling, “Did you ever have the feelin’ that you wanted to go, but still have the feelin’ that you wanted to stay. Go or stay, stay or go, start to go again and change your mind again!”

Elon Musk’s up and down attitude towards Bitcoin echoes Jimmy’s “did you ever have the feelin'” predicament. The Tesla CEO famously bought $1.5 billion in the signature cryptocurrency early this year, igniting a giant rally, then backtracked in May by tweeting that Tesla would no longer accept Bitcoin as payment for its EVs due to its extensive carbon footprint, sparking a selloff. Then, it seemed that Musk might be abandoning Bitcoin in early June, when he issued a meme about a couple breaking up––sporting a Bitcoin hashtag and broken heart emoji. Two weeks later, the auto industry’s premier showman appeared to shelve his stetson and sit back down at the keyboard by tweeting that Tesla would resume accepting Bitcoin “When there’s confirmation of reasonable (around 50%) clean energy usage by miners.”

Now, a new report is chronicling how the carbon used to mine Tesla’s Bitcoin is negating a significant part of the emissions saved by its EVs. It provides numbers documenting that Tesla’ green goals and enthusiasm for Bitcoin contradict one another. Its findings should give Musk good reason to get the feelin’ to go.

The study is titled “The True Costs of Digital Currencies: Exploring Impact Beyond Energy Use,” published in the Elsevier journal One Earth. The authors are economist Alex de Vries of the central bank of the Netherlands, who also runs the site Digiconomist that tracks Bitcoin’s energy footprint, Ulrich Gallersdorfer and Christian Stoll of the University of Munich, and Lena Klaassen of MIT. The report offers a broad view of Bitcoin’s shortcomings in all three categories of ESG, energy, social impact, and governance. But I’ll focus on how the authors, for the first time, quantify the share of Bitcoin’s emissions attributable to Tesla’s holdings.

They note that Bitcoin mining consumes almost as much energy as all the the world’s data centers combined, and produces CO2 emissions that match the carbon footprint for the city of London. The paper estimates that Bitcoin generates 90.2 million metric tons of carbon gases per year. That volume exceeds by almost two times the total annual reductions achieved by the replacement of gasoline-burning cars by electric vehicles, as calculated by the International Energy Agency.

To arrive at Tesla’s “share” of Bitcoin’s emissions, the authors adapt the Greenhouse Gas (GGH) Protocol that determines the portion of a company’s carbon footprint attributable to its equity investments in other enterprises that emit hydrocarbons. Say company A is a passive investor in company B and has no power to determine B’s energy policies. If A owns 20% of B, the GGH system simply allocates one-fifth of B’s carbon footprint to A. Obviously, Bitcoin isn’t a company, but for Tesla, it’s clearly an investment. Hence, it makes sense to allot Tesla the proportion of Bitcoin’s carbon emissions that matches its ownership share in all Bitcoin.

The authors reckon that Tesla acquired 43,345 coins during its big buy at the start of 2021, and still holds 39,092, following the famous sale of 10% of their stash later in Q1. (My own estimate of their current haul is a bit higher at around 41,000). Miners have minted almost 19 million Bitcoin, but around 20% of the coins have been lost. So of the roughly 15 million in circulation, Tesla owns about 0.26%. Its “share” of Bitcoin’s annual carbon emissions is one-quarter-of-one-percent of that 90.2 million Mt., or 0.24 million Mt.

So how does that number compare with the emissions that Tesla claims its EVs are saving each year? Tesla doesn’t report the numbers by year, but does provide other data that point to an estimate. I arrived at the following approximation with the assistance of co-author de Vries. In an environmental impact report from April of 2019, Tesla stated that it had saved 4 million metric tons (Mt.) of CO2 over its history through that date. It recently posted on its website that the current number’s risen to over 16 million Mt. Hence, over the past two years, Tesla reckons that it’s reduced emissions by a total of around more than 12 million Mt., or 6 million Mt. annually. So we’ll put the reduction in achieved in 2020 at half the two-year total or 6 Mt.

Conclusion: The emissions that produced the almost 40,000 coins on Tesla’s balance sheet equaled about 4% of the gases that its cars kept from polluting the atmosphere. In 2020, Tesla delivered 181,000 vehicles. You could argue that its romance with Bitcoin erased the environmental benefits from more than 7,000 of those EVs. De Vries says that if a buyer purchases a Tesla by making payments over 48 months in Bitcoin, the carbon footprint of the Bitcoin transactions would exceed the total lifetime emissions savings from driving an EV instead of a gasoline-powered car.

Of course, we don’t know if Musk will keep buying Bitcoin for Tesla’s treasury, or resume accepting coins for its Models S, X, Y, and 3. But every coin that Musk buys or collects was mined by putting gases in the air that Tesla’s EVs are taking out. Every coin in its treasury undermines its image as the ultimate green champion. No wonder that when it comes to Bitcoin, Elon Musk doesn’t know if he’s coming or going.

Subscribe to Fortune Daily to get essential business stories straight to your inbox each morning.