Bitcoin mining is suddenly one of the most profitable businesses on the planet

Obviously, it’s this summer’s sweeping clampdown in China that has greatly swelled profits for miners outside of the world’s second-largest economy. You don’t see many industries where, virtually overnight, the government authorities banish half the world’s competitors. The miners who never stopped producing in Quebec, Texas, Kazakhstan, and Malaysia are harvesting a giant windfall. But just how big is that bonanza, and how much money are they currently making? To be sure, industry data is hard to come by. Some miners are clearly much more profitable than others. That’s because rates for electricity, the principal expense, vary widely from locale to locale, and miners loaded with the newest, most powerful machines earn more than those running on “legacy” gear that’s as little as two years old. Still, we know the size of the industry’s revenues, and we can get a reasonable estimate of overall electricity costs. Those figures can give us a good view of the overall industry’s newly exploding margins.

How big is the global Bitcoin business?

The industry’s overall size is easy to calculate. The world’s producers earn 6.25 coins every 10 minutes, or 328,500 a year. At the price of roughly $39,000 at mid-afternoon on Aug. 4, those “block rewards” are running at $12.8 billion a year. Miners also earn transaction fees each time they win a new batch of coins, generally ranging between 5% and 10% of the reward and paid in Bitcoin. Using the lower number, we get total revenues of $13.5 billion. By the way, that figure exceeds the anticipated 2021 revenues from gold and silver mining by around 5%. (Bitcoin’s carbon footprint is also about the same size as emissions from mining the precious metals.)

To get the first rough look at the industry’s profitability, it’s crucial to calculate what miners are paying for their biggest nut, electricity. Alex de Vries, a Dutch economist whose website Digiconomist tracks Bitcoin’s carbon footprint, estimates the producers’ average cost at 5 cents per kilowatt-hour (kWh). That’s the blend of prices ranging from as low as 2 cents in Argentina and 3 cents in Kazakhstan to 8 cents and above in the U.S. Prior to the China shutdown, de Vries estimates, the industry was spending about $7.2 billion on electricity. We don’t know exactly how China’s electrical costs compared with the average in the rest of the world. But we’ll assume that the total energy for China’s miners more or less corresponded to its 50% slice of all Bitcoin production. In that case, today’s active miners are paying around $3.6 billion a year for power (half the $7.2 billion total before the lockdown).

Here’s where the China upheaval delivered such a prize to the rest of the world’s miners. Over a few weeks in June and July, nearly 50% of the industry’s machines, heavily concentrated in Xinjiang province in the north, and Sichuan and Yunnan in the south, went offline. The Bitcoin algorithm aligns the difficulty of winning awards to the global “hash rate” or network’s total computing power. Miners win coins in direct proportion to the overall hash rate they generate. When China went offline, half the world’s computing power vanished. As a result, the producers still cranking doubled their market share. Instead of winning half the Bitcoin released each day, they began winning all the coins. In a flash, non-Chinese miners— without changing anything, and without adding power generation, machines, or people—doubled their winnings.

The upshot: Instead of capturing half the global revenues, the “rest of the world” miners are getting it all. Suddenly, they’ve gone from earning newly issued Bitcoin worth $6.75 billion a year to devouring the entire pie of $13.5 billion. But their costs of labor, rent, insurance, and the big-E, electricity, are fixed. They’re still paying the same, $3.6 billion for power, for example, while their revenues doubled. Before the China storm, they were posting gross profits after deducting electricity costs of $3.1 billion ($6.7 billion in revenues less $3.6 billion for power), or 46%. Now, they’re making $9.8 billion in gross profit on newly doubled revenues of $13.5 billion, for a margin of 73%.

For top producers, operating margins are heading to over 70%

Of course, the miners shoulder sundry smaller operating costs besides by far the largest one, electricity. Those include rental on the buildings housing the computers, corporate overhead, and depreciation that consists of expensing the cost of their equipment over its useful life. Starting with the data on overall revenues and electricity costs, let’s estimate operating margins both for the entire industry, and for the most productive miners.

An excellent case study in how the China crunch benefits a big, highly efficient producer is the roaring liftoff at Bitfarms. Founded in 2017, Bitfarms runs 100% on clean energy, chiefly hydropower drawn from the raging rivers of its native Quebec. Bitfarms was already robustly profitable on an operating basis before the exodus from China. In February, it mined a total of 178 Bitcoin; that month, the price per coin had started on its moonshot and averaged $44,000. According to company reports, Bitfarms in February generated roughly $7.8 million in “mining profits,” defined as revenues less the basic production costs of electricity and “infrastructure.” The latter includes the rental on the warehouses holding its computers, and maintaining the machines. Bitfarms disclosed that its mining costs for each Bitcoin earned amounted to $8,400. The confluence of Bitcoin’s bounding price and ultralow costs delivered a mining margin for the month of 80%.

That elevated number signals that Bitfarms is a lot more productive than most miners. Last year, Bitfarms reported that it’s paying an average of 4 cents per kWh for electricity, 20% below the estimated global average. The flight of its Chinese competitors took Bitfarms’ profitability to a whole new level. In July it mined around 391 coins, double the total in February. That’s 13.5 coins a day versus roughly seven just five months earlier. Even at an average price almost one-fifth lower of $35,000, Bitfarms by Fortune’s estimate booked over $14 million in mining profits, 80% more than in February. Put simply, Bitfarms hugely raised revenues by winning twice as many coins, while its costs remained flat. Result: Its mining margin rose from 80% to almost 90%.

So what’s the run rate for Bitfarms’ overall operating margins, encompassing not just mining costs but the other relevant expenses? At midday on Aug. 4, Bitcoin’s price had rebounded to over $39,000. At its current output of 400 coins per month, Bitfarms’ run-rate for revenue is approximately $190 million a year. Its mining costs of electricity and infrastructure remain around $20 million a year. By winning twice as many coins without paying more for power and facilities, it has cut mining costs per coin in half, from $8,400 to $4,200. In the first quarter, its annualized depreciation and amortization was $12 million and SG&A $11 million. All told, its yearly operating expenses appear to be in the $43 million range. If so, its operating profits are clicking at around $150 million a year. That gives Bitfarms an operating margin in the 75% range.

By comparison, in their most recent fiscal years, LVMH posted an operating margin of 18%; Apple 27%; ADP 32%; Microsoft 37%; and Amgen[/hotlink] 38%.

How profitable is the overall Bitcoin industry?

From the Bitfarms example, we can conclude that the most efficient miners are feasting at operating margins of 70% and even higher. Those players generally benefit from the combination of operating in venues with extra-low electricity prices and deploying the most powerful computers. Every 15 months or so, computer manufacturers—the largest by far is Bitmain of China—introduce a new suite of specialized ASIC computers that are more potent than the previous line. The most efficient miners are constantly replacing their older machines with the updated ones to gain market share with the same number of computers. For example, the new Bitmain S-19 generates almost eight times the hashes per second as the older S-9 and uses half the electricity per unit of power. “It’s a constant race to keep moving up to the newest, most powerful machines at faster and faster intervals,” says de Vries. “The miners have only a short time to earn their money back on those machines.”

According to de Vries, miners even in countries with low energy prices such as Kazakhstan can carry higher than average operating costs if they’re using older computers. That helps explain why the industry as a whole has a margin after subtracting only electricity costs of 73%, below Bitfarms’ figure including all of its operating expenses. If we assume that most miners have SG&A, depreciation, and rental and maintenance about equal to Bitfarms’ numbers as a share of sales, those expenses would equal an additional 17% of revenues. Deducting those 17 points from the margin of 73% after electricity costs, it appears that the Bitcoin universe is running at operating margins of around 55%. That still beats the profitability for virtually any other big industry.

This golden period won’t last

Bitcoin mining’s fantastic profitability arises from the convergence of three forces. The first is the surge in prices to multiples of their level just two years ago. Second, a shortage of semiconductors blocked both aspiring newcomers and entrenched players from buying computers to capitalize on Bitcoin’s bull run. That bottleneck left the incumbents with an oversize market share that rendered the field highly lucrative even before the Chinese exit. The third, of course, was that sudden shuttering of half the industry.

Over time, Bitcoin’s big profits will attract a flood of competitors, whittling those humongous profits to slender profits. The Chinese are desperate to relocate and resume mining, because the faster they get up and running, the more money they’ll make. The return of the Chinese alone would return margins to where they were before the great departure, assuming prices remain at today’s levels—making the industry half as profitable as it is today.

As the global hash rate goes back to where it was earlier this year and keeps rising from there, miners will get fewer and fewer coins using the same amount of electricity. They’ll have to buy costlier and costlier machines to stay competitive. Eventually, the expense of winning a Bitcoin, plus the small profit required to keep miners in the game, will rise to equal the price of a Bitcoin.

The bull case is the one that’s prevailed all during the boom times for Bitcoin: Its price will keep racing faster than existing players and new competitors can add new machines. But prices can also go the other way, shortening the time it will take production costs to more or less equal prices. In the long run, the business of Bitcoin will resemble the dull, grinding ordeal of scouring the earth for copper or silver or gold ore. The Bitcoin miners are incredibly profitable in large part because they’re incredibly lucky. But their big strike will start a gold rush that will leave a lot less gold to go around.