Even before the Chinese government ordered a total shutdown of Bitcoin production and trading, the business of “mining” the world’s top cryptocurrency by market cap was yielding astounding profits. Sprawling plants the size of regional power stations, packed with specialized ASIC mainframes whirring 24/7, were generating margins that dwarfed the rewards of unearthing rival “stores of value” gold and silver.
Even after prices dropped from almost $65,000 a coin in April to around $35,000 in mid-June, business was so good that the big players were placing record orders for the newest generation of “rigs.” Only a global shortage of semiconductors blocked miners from vastly expanding their computing power to land bigger and bigger awards of newly minted Bitcoin.
The crackdown in China is a brutal blow to that nation’s miners, who reportedly were generating around half of the world’s Bitcoin output. But at least for now, it’s a huge windfall for competitors in the U.S., Canada, and Russia, who are operating at the same capacity today as before the hammer fell in the world’s second-largest economy.
The top mining nation shuts it down
Though Beijing’s hostility has been building for months, the death sentence came in mid-June. Most of China’s production flowed from miners who split the year between northern Xinjiang province for the fall and winter months, and Sichuan and Yunnan in China’s southern tier for the spring and summer, shuttling their equipment back and forth 2,000 miles or more in tractor trailers. The annual migration was a quest for cheap electricity: The miners ran on bargain coal in Xinjiang, then got an even better deal on excess hydroelectric power fed by the thunderous monsoons in Sichuan and Yunnan.
On June 19, Sichuan and Yunnan provinces ordered all miners to depart. A few days later, the central government imposed what amounted to a nationwide ban by commanding Chinese banks to end all dealings with Bitcoin producers.
Within a week, more than nine in 10 Chinese miners shuttered their farms, and started scouting for promising places to relocate, from Kazakhstan to Texas. From the peak on April 19, the global computational power fell by approximately one-half.
Surprisingly, the upheaval didn’t much ruffle Bitcoin’s price, which has toggled since then between $31,000 and $35,000. But the shock will cause two principal shifts. First, profits for the non-Chinese miners are likely to as much as double from already lofty levels. For how long depends on how fast the displaced find new homes—and it could take as long as two years or more for the approximately 4 million mining machines leaving China to resurface in new locales.
Second, the exodus from Sichuan and Yunnan, the places hosting all but a small portion of the world’s green crypto production, means that fossil fuels are now powering a much bigger share of the world’s Bitcoin output than before the shutdown. The Chinese miners are already shipping their gear to parts of the world where production is most active, notably such U.S. states as Texas and Tennessee, and a roster of welcoming nations including Kazakhstan, Russia, and Malaysia. These are all venues where most of the Bitcoin, along with the vast bulk of all the electricity consumed by businesses and households, is generated by coal and natural gas.
The great migration from China is just one of two major forces threatening to worsen Bitcoin’s carbon footprint, estimated before the shutdown to equal that of Argentina. Because of the China earthquake, miners in the rest of the world are making much more money, so they harbor loads of cash and a yen to spend it on towering racks of the newest, most powerful ASIC computers. These lucky incumbents are mainly relying on fossil fuels. They will likely be burning a lot more coal and gas to fill the void left by the sudden shuttering of what was the industry’s engine room.
Of course, a strong trend toward renewables could take hold. Hydropower in Canada and even nuclear power in Ohio offer promising opportunities. But for now, the best bet is that big, looming increases in output come in the places producing most of the coins today. And they are mainly chugging along not on wind, solar, and hydro, but on feedstocks that spew carbon dioxide.
If Bitcoin takes another moonshot from $32,000, the rush to install more computers, devour more electricity, and pump out more CO2 will only accelerate. “At $28,000 per coin, electrical usage could rise by 30% above the level prior to the China shutdown, and the miners would still make good money,” says Alex de Vries, an economist whose Digiconomist website tracks Bitcoin’s energy usage. “If it goes back to $65,000, the miners would double their power usage from before China unplugged.”
It’s hard to find a better window into what lies ahead than studying the largest publicly traded, “all green” Bitcoin producer in North America, Bitfarms of Canada. Bitfarms runs 100% on hydroelectric power. In a single example, it illustrates the fat rewards of minting Bitcoin before the China typhoon; the still bigger profits brought by the sudden uprooting of half your competitors; and the difficulty of making renewable energy the dominant power source for Bitcoin manufacturing.
Bitcoin’s automatic stabilizer
Before we get to Bitfarms, let’s examine how China’s going dark changes the way the Bitcoin network dispenses new coins, and how that reset benefits the miners running at full tilt.
The Bitcoin network is designed to dispense 6.25 coins every 10 minutes, or 900 per day. Miners vie for awards in a kind of giant guessing game or lottery based on running “hashes,” or multi-digit codes generated by their ASIC rigs. The one miner that hits the code assigned to the next batch of 6.25 Bitcoins gets the full award.
The winner receives not just the coins, but fees for transfers and purchases that the miner authenticates by signing the “block” containing those transactions. Those fees provide additional income that on average, comes to roughly 10% of the Bitcoin price. The fees are paid in Bitcoin by the people and companies performing the transactions. So the total “block awards” encompass the 900 Bitcoins per day, plus another 100 paid in fees, for a total of around 1,000 coins.
The algorithms governing the Bitcoin network impose a kind of automatic stabilizer. The system adjusts the difficulty of winning awards to the total capacity, or “hash rate,” in the network. Rising prices lure miners to buy more and more machines, raising the total systemwide hash rate. As that happens, the network mandates proportionately more hashes, or computer-generated guesses, to win the same amount of Bitcoins. If the hash rate doubles, the number of Bitcoins awarded doesn’t double—the system is engineered to keep the number constant at 6.25 every 10 minutes. Instead, the network raises the bar so that the average miner needs to generate twice as many hashes to garner the same number of Bitcoins.
The way the model is designed, the only way to win more coins is to gain a bigger share of the network’s total computing power—in other words, boost your portion of the overall hash rate versus competitors. But if the price of each Bitcoin waxes a lot faster than the rise in the hash rate, the miners who don’t add any machines make a lot more money, even though their share of the total market declines. And those that expand do even better, because the returns on their investment in those new banks of computers is extremely high.
That’s precisely how the market has worked since the start of the pandemic. As the price leaped 10-fold from March 2020 through April 2021, the miners did increase the overall hash rate, but the profits from mining rose much faster. To capitalize on the booming market, miners normally would have installed far more capacity than they actually added. Indeed, orders for the newest ASIC machines soared. But the semiconductor shortage and supply chain disruptions caused big backlogs, keeping the overall hash rate much lower than if the machines had arrived. “The hash rate about doubled in the past year. But if all the machines on order had been installed, it would have gone up a lot more,” says de Vries.
The upshot: Miners who were solidly ensconced last March benefited two ways. First, their product’s price mushroomed. At $32,500 on July 8, Bitcoin is still at many multiples of its value just before the pandemic struck. Second, the chip crunch effectively “grandfathered” those entrenched players by giving them a much higher market share than if rivals had been free to add all the new capacity they craved. The confluence of those warm currents appeared to make the past 18 months the golden age of mining. Now, the upheaval in China spells even richer days ahead.
Making it easier to win new coins
Just prior to the great unplugging in China, the total capacity of the network peaked at 191 exahash per second in mid-April. That’s a rate of 1 quintillion codes per second, or 1 followed by 18 zeros. At the lowest point the week of June 27, the figure fell by half, to around 80 exahash. By July 15, it had stabilized in the 100-exahash range. In the space of two weeks, the Bitcoin network’s total computational power shrank from the summit by around one-half.
Here’s an example illustrating how the China exodus is leading to greater profits for overseas rivals. Before the capacity crash, a phantom producer we’ll call Major Miner generating 1.7 exahash would have held 1% of the market. That share would deliver 10 Bitcoins of the 1,000 coins won in block awards each day. The network’s architecture dictates that because capacity dropped 50%, the hash rate needed to secure an award also falls by half.
But that adjustment doesn’t happen instantaneously. It usually takes about two weeks for the system to reset. This time, the drop was so severe that the adjustment took a week longer. Because of that built-in delay, the network made an interim shift. For those three weeks, the protocol downshifted to issue the regular 6.25 Bitcoin at longer intervals. Instead of every 10 minutes, the release times between mid-June and the first few days in July slowed to an average of 14 minutes, a delay of 40%.
As a result, the China shutdown, even in the interim period, made mining much more profitable for players outside that country. The network issued far fewer coins than usual, but the proportion of hashes generated by incumbent miners rose faster than the slowdown in new coins. As we’ll see, miners such as Bitfarms are already feasting. And the biggest profits are yet to come.
On July 3, the automatic stabilizer kicked in, lowering the hash rate needed to capture new coins by 28%. A second adjustment is coming on July 17, with more to follow, until the computing power needed to win a batch of Bitcoin falls by one-half. Miners that owned 1% of the computational power good for 10 coins a day will soon be banking 20 without adding any new machines. The same number of electronic guesses will generate twice as many Bitcoins. It’s only in the wild world of Bitcoin that half your competitors retreat overnight, so that you can change nothing and double your profits.
For Bitfarms, a great business just got even better
Bitfarms was thriving before the China lockdown—but now it’s poised to become far more profitable. Its before-and-after numbers are a case study in where the industry’s coming from, and the extra riches ahead.
Founded in 2017, Bitfarms operates five plants, with a sixth underway, that run on electricity generated by the raging rivers of Quebec. The province, in fact, supplies over 99% of all its energy for consumers and businesses via hydro. Bitfarms went public in 2019 on the Toronto Stock Exchange, and in June it won a listing on the Nasdaq. At a market cap of $600 million, it ranks in the top half-dozen of all publicly traded Bitcoin miners. Since late last year, its stock price has vaulted sixfold, to $3.22 per share. Starting in January, Bitfarms has been parking 95% of the Bitcoin it collects into its treasury.
According to Bitfarms’ public disclosures, its “mining costs” per Bitcoin won in Q1 2021, consisting of cost of supplies, labor, and electricity, was $8,500. It benefits greatly from the ultralow cost of hydropower. Last year it posted heady gross mining profits, excluding overhead, of 55%. But that was just a warmup for 2021.
In Q1, Bitfarms mined 598 Bitcoins to amass $28 million in revenue. Its production cost was just $6 million, meaning that its gross margins jumped to 79%. Clearly, its low cost structure made Bitfarms profitable at prices in 2020 that were much lower than today’s, and the cryptocurrency’s big run so far this year is stoking an earnings blowout.
Consider the amazing ascent in daily profits. In March of 2020, Bitfarms was booking just $35,000 every 24 hours. By December, the figure had advanced to $90,000, rising to $163,000 in January, $238,000 in February, and $330,000 in March. That month, it was making 85¢ in gross profit for every dollar worth of Bitcoin it mined.
Of course, March and April marked peak Bitcoin, as its price ranged between $50,000 and $64,000. But although Bitcoin’s value has dropped by half, Bitfarms is pocketing just as much profit as at the summit—because it’s grabbing around twice as many coins. Its June “production update” doesn’t disclose earnings but shows the number of Bitcoins won per month. In June, Bitfarms gained 265 Bitcoins, one-third more than in January, and at its current pace, it will reach 415 in July.
As the production update explained, “On July 3, the Bitcoin network experienced the largest difficulty drop in history due to recent macro developments in China. This has resulted in Bitfarms producing significantly higher quantities of Bitcoin at lower cost per Bitcoin.” Founder and CEO Emiliano Grodzki noted that “Bitfarms has nearly doubled its market share.” From the start of 2021, its winnings have doubled from seven to 14 coins per day.
Bitfarms’ operating costs are almost entirely fixed, since electricity is by far the biggest expense. So it’s now minting twice as much product at the same cost. By Fortune’s estimate, its expense per Bitcoin suddenly dropped from $8,500 to around $4,500. “Our revenue for the same operating cost just doubled,” chief mining officer Ben Gagnon tells Fortune.
Today, Bitfarms holds 1.42 exahash of installed production. When the total for all mining stood at 191 exahash before the China shock, it controlled around 0.74% of the market.
But since the total network has now shrunk to 90 exahash, Bitfarms’ portion of the pool is coiled for a spring to 1.5%. With the next reset, its daily Bitcoin production should rise from 13 to 14 per day. At current prices, that’s good for daily revenue of as much as $500,000, though half as much as in the bountiful month of March when Bitcoin itself was more valuable.
Chinese miners will return, but the challenge is huge
How long the bounty will last depends on how fast the Chinese miners find new homes. “They need to relocate power equivalent to that produced in all of Belgium,” says de Vries. Gagnon of Bitfarms predicts that it will take two and a half years for the Chinese to build new factories abroad. He notes that before the shock, those miners’ biggest problem was getting new machines, mostly produced in their home country by Bitmain, the world’s largest supplier. “The industry was ordering 100,000 ASIC computers a month because the prices and profits were so high,” says Gagnon. “Fifty-five to 70% of the customers were Chinese. Now, the ‘shortage’ issue has totally shifted, from machines to electricity. The new bottleneck is the Chinese miners’ search for scarce power.”
Gagnon, who has worked extensively in China, explains how shepherding a project from blueprint to churning hashes is 10 times more arduous and time consuming in a place like Texas than on their home turf. In China, says Gagnon, “the miners will drive up to a coal plant or hydro dam. They’ll have a couple of drinks with the plant’s owners, then dinner, and do a handshake deal. The next day, they’ll arrive with their equipment.” The electrical contractors and other workers will toil 18 hours a day, seven days a week, erecting the plant holding the computers. “They’ll sleep on-site,” says Gagnon. “That doesn’t happen in the rest of the world.”
What took eight weeks in China will likely take the Bitcoin migrant miners eight months to accomplish in Canada, says Gagnon. And in the U.S. or Europe, the Chinese miners would need to secure permits, negotiate agreements with utilities, and hire contractors who work at a much slower pace. “The infrastructure isn’t there, and the agreements aren’t there,” says Gagnon. The displaced can import their computers, but not their electrical gear. “They need to install switchgear panels, breakers, transformers, and power distribution units (PDUs),” he says. “That equipment has different certification standards in the U.S. and Europe than in China. They can’t be reused like the computers. All of that equipment needs to be replaced.” The COVID crisis has stretched supply lines for these items, causing backlogs. Delays in getting those electrical essentials could greatly extend the timeline for rebooting abroad.
Most of all, says Gagnon, the challenge will be finding sufficient power. “China is the No. 1 electricity producer in the world, far bigger than the U.S.,” he says, noting that the miners there benefited from the nation’s giant excess power generation. That surplus doesn’t exist in many places, says Gagnon. He adds that the challenge isn’t just relocating more than half of the world’s existing machines. The Chinese also need to find a place for the hundreds of thousands of computers they have already ordered that are stacking up in storage.
Bitcoin’s carbon footprint will likely keep growing
As the Chinese seek new havens, and the miners still cranking expand to profit from still-strong prices, how likely is it that energy usage will shift from fossil fuels to renewables? It’s important to understand that, believe it or not, China’s dominance has been waning in recent years. De Vries reckons that as recently as September 2019, the No. 2 economy accounted for 75% of the entire network. But many Chinese miners ventured abroad, perhaps fearing that Beijing and the provinces would eventually crack down. In less than two years, the U.S. share rose from 4% to 17%, Iraq’s from 2% to 6%, and Kazakhstan’s from 1% to 5%.
That trend shrank Bitcoin’s green footprint before the coup de grace. That’s because the quantities mined via hydro in Sichuan and Yunnan fell along with China’s overall slice of the world’s computational power. In April 2021, Yunnan and Sichuan accounted for just 29% of the Bitcoin network, down from 44% in April of 2020, according to a study by de Vries. An estimate by Cambridge University put green production worldwide at about one-third of the total, meaning Sichuan and Yunnan were the main show. The role of nations that generate a large proportion of their energy through renewables is relatively minor. Canada contributes just 3% of the network’s computational power, while Europe’s hydro capital Norway provides a fraction of 1%.
So if Bitcoin keeps getting mined where it’s being mined now, its carbon footprint is likely to grow. It’s possible that won’t be the case. Miners could flock to places that offer excess energy furnished by renewables. For example, Washington State enjoys a big surplus of hydro. Still, miners would need to obtain separate permits from individual counties and the federal government to tap the grids. It’s not at all clear that the Chinese could secure those regulatory approvals. Another option is an overlooked source that doesn’t emit CO2, but isn’t exactly regarded as “green.” Mining specialist Standard Power just signed a five-year deal in Ohio to produce Bitcoin using nuclear power.
In Canada, the chances of replicating what Bitfarms has created are slim. The Quebec government has capped the amount of hydro that can be used for Bitcoin mining, and it has allocated over half that capacity to Bitfarms. Hence, the potential for new competitors in Quebec is severely limited. So far, wind and solar have gained little traction. “Wind and solar are particularly bad matches for crypto,” says de Vries. “Miners buy expensive machines that need to be replaced every 18 months by newer, faster machines. To be profitable, miners need to run those computers 24/7. They can’t be shut down waiting for the sun to shine or the wind to blow.”
Right now, the only thing likely to stop Bitcoin mining’s resurgence from the Chinese setback is a collapse in its value. Even though mining gets easier as the price drops, if it goes low enough, the cost of electricity and equipment will make the enterprise unprofitable. It’s a scenario that the believers who think crypto will change the world judge preposterous. Here’s the puzzle for the zealots. The only way to ensure that Bitcoin’s carbon footprint shrinks, sans government lockdowns, is for the price to dive. Just ask Elon Musk. Being green and hugging Bitcoin at the same time is the toughest of tough acts.
A version of this article appears in the August/September 2021 issue of Fortune with the headline, “Can Bitcoin go green?”
Subscribe to Fortune Daily to get essential business stories straight to your inbox each morning.