This post was originally published on this site.
Over the weekend, bitcoin miner Cango sold 4,451 bitcoin for around $305 million. According to a company press release, the sale was completed to partially pay down a bitcoin-collateralized loan. Additionally, the sale is said to strengthen the company’s balance sheet, reduce financial leverage, and, notably, provide the capacity for an expansion into AI infrastructure.
Notably, the sale comes after the bitcoin price was basically cut in half from its October all-time high when it hit around $125,000; however, Cango also made it clear that they’re still committed to their bitcoin mining operations, despite the additional expansion into AI. The company says the transaction was settled in Tether’s USDT stablecoin.
Cango is far from the first bitcoin mining company to get involved with the AI data center industry. While the chips used in bitcoin mining are made specifically for hashing the SHA-256 algorithm and are not applicable to the AI industry, the basic, real-world infrastructure built around those machines is very similar to what is needed for AI, especially in terms of access to a large amount of cheap electricity that can be used to power any computer hardware all day long.
Some of the other bitcoin mining companies that have already expanded into the realm of AI include IREN, Core Scientific, and Riot Platforms. And yes, some bitcoin mining companies have indeed abandoned the crypto asset completely in search of AI-focused profits.
It’s sometimes difficult to tell when a publicly-traded company is pivoting to a new technology for business reasons or to market its stock to retail buyers. The recent mentions of AI in press releases from all kinds of different companies traded on the stock market are reminiscent of the similar “blockchain” phenomenon of 2017 and 2018, where everyone from iced tea companies to camera manufacturer Kodak was touting the revolutionary aspects of blockchain technology that would revitalize their companies. There are undoubtedly at least a handful of companies that were pivoting to blockchain back then that are now touting their new AI capabilities. But as already mentioned, there are also clear reasons an expansion into AI is likely to make business sense for those already involved with bitcoin mining.
Bitcoin’s recent price decline has had a direct impact on the very asset that miners generate with their hardware and associated energy costs, although the recent snowstorm that ran through much of the country turned out to be quite profitable for those who have curtailment deals for selling electricity back to the grid when everyone is staying home and using their heat. Meanwhile, other bitcoin mining startups are looking at ways to integrate machines into home heating equipment as a way to offset the costs of climate control.
Of course, the other aspect of this sale that is interesting is that Cango is paying down a leveraged position on Bitcoin at a time when the price has been in a downward spiral. Many companies have taken on bitcoin as a long-term reserve asset in their treasuries over the past couple of years, so that the sale of around $305 million worth of bitcoin from such a treasury is notable from that perspective.
There are worries that deleveraging events with some of the larger digital asset treasury (DAT) companies that leverage debt in their strategies, namely bitcoin treasury company Strategy, could lead to a downward death spiral for the bitcoin price; however, Strategy recently stated that the price would have to drop to $8,000 for real concerns regarding their balance sheet to pop up. That said, Strategy saw massive unrealized losses of $17 billion in the last quarter of 2025, and a separate DAT focused on Ethereum is already dealing with $7.5 billion in unrealized losses less than a year since its original launch.
Some DAT executives at the recent Digital Assets at Duke event have indicated they can weather the recent price turmoil, as their success or failure will be measured over much longer timeframes. But it’s also worth remembering that not all DAT companies are created and managed equally.