Article by: Eric, Foresight News
Cangu, an automotive financial technology service platform founded in August 2010, raised several hundred million yuan in Series B funding in 2018 led by Tencent and Taikang Life, and listed on the NYSE in July of the same year.
Bitmain, a cryptocurrency mining hardware chip and complete machine design, manufacturing, and mining pool operator founded in 2013, completed three rounds of funding totaling nearly $800 million between 2017 and 2018, with a valuation once reaching $15 billion. It failed to go public on the Hong Kong Stock Exchange at the end of Q3 2018.
From a business perspective, these two seemingly unrelated companies have quietly become intertwined by fate by the end of 2024.
From Old Finance to New Finance
Sixteen years ago, in the summer, Cangu started in Shanghai, building a nationwide network of automotive finance channels covering lower-tier cities and counties, entering the market through “auto loan facilitation.” Founder Zhang Xiaojun accumulated over ten years of experience at SAIC Group and resigned to establish Cangu after serving as Deputy General Manager and Director of SAIC General Automotive Finance.
At 43, venturing into entrepreneurship, he shows signs of a late bloomer. Relying on his extensive industry experience, Cangu initially targeted the long-tail market of solving car loan difficulties for residents in small cities, quickly earning its first pot of gold. Later, Cangu shifted its focus to auto sales and the aftermarket, establishing a comprehensive automotive transaction service platform on top of a stable core, connecting dealers, financial institutions, and consumers, forming an integrated service system of finance, sales, and after-sales.
In 2018, Uxin, Guazi, and Renrenche were aggressively advertising and engaging in price wars, while Cangu reported revenues of 430 million yuan in 2016 and 1.05 billion yuan in 2017, with net profits of 134 million yuan and 349 million yuan, respectively. However, in terms of fame, Juliet Ye, the PR & IR director hired in 2018, said that Cangu was almost “unknown outside the industry.”
This low-profile company, which seems to be somewhat over the top, had connected 37,700 registered traders, 11 third-party financial institutions, and 29 other industry participants including OEMs, online advertising platforms, and insurance brokers by the time of its IPO. Since its founding, it has served over 730,000 car buyers. In 2017, the auto transactions facilitated by Cangu accounted for 3% of all auto finance transactions in China that year, ranking first.
Looking back, 2018 appears to be Cangu’s peak. Although it experienced a small spring in early 2021 with the IPO of its invested company, Li Auto, its performance continued to decline. By 2022, Cangu’s market value had fallen from $11 to $0.5 per share, with a loss exceeding 1.1 billion yuan that year. The reason was that in 2020, Cangu shifted its focus from finance to auto trading, but was hit by the expansion of new energy vehicle direct sales models and a rapid decline in traditional car sales.
Seeing no way out, Cangu finally decided to pivot at the end of 2024, or at least, as they put it, to “transform.” The direction of this transformation is Bitcoin mining.
In November 2024, Cangu officially announced its shift to a Bitcoin mining company, starting with a $256 million cash purchase of mining machines from Bitmain, with a total hash rate of 32 EH/s. It also plans to invest an additional $144 million by issuing new shares to buy second-hand mining machines with a total hash rate of 18 EH/s from sellers like Golden TechGen Limited. A total of $400 million, Cangu immediately became one of the top three companies globally in hash rate. The second-ranked CleanSpark recently achieved 50 EH/s at the end of last year.
Previously bankrupt EV brands like GAC Aion and JEV have turned to Middle Eastern crypto companies, and Faraday Future’s CEO Jia will also announced a move into Web3. Some jokes say: the end of struggling car companies is cryptocurrency. For Cangu, starting in auto finance and now stepping into a new financial field is not exactly a cross-industry move.
Many US-listed companies have transformed into Bitcoin mining firms, but Cangu’s bold move to spend $400 million on mining machines right out of the gate is unprecedented. Such a high-stakes gamble, while astonishing, also raises questions: is this sudden shift too smooth, too perfect?
Transformation or Shell Listing?
Through careful analysis of subsequent actions, we can almost confirm that Cangu’s sudden transformation was actually a shell for Bitmain’s long-planned listing.
In 2018, Bitmain submitted a prospectus to the Hong Kong Stock Exchange, planning a valuation of up to $50 billion. According to the prospectus, Bitmain’s net profit in 2017 exceeded $1.1 billion, with an estimated $2.2 billion in 2018. Among many unprofitable new economy companies, Bitmain excited the capital markets.
However, in 2018, coinciding with the crypto bear market, Bitcoin prices plummeted, sharply reducing demand for mining machines, causing volatile performance. Investors and regulators worried about business sustainability and regulatory uncertainty across multiple countries. The HKEX also expressed caution toward the crypto industry. Additionally, Bitmain’s plan to pivot to AI chips became an obstacle; HKEX President Li Xiaojia publicly stated that if a company made big money from mining hardware but then suddenly emphasized AI chips without substantial results, its original model lacked sustainability.
Most critically, in early 2019, Bitmain co-founders Wu Jihan and Jihan Zhang engaged in a fierce power struggle, leading to a split. Wu Jihan left to establish Bitdeer, causing Bitmain’s valuation to plummet to around $5 billion, losing its IPO prospects entirely. Rumors of a US IPO attempt surfaced afterward, but ultimately, it was unsuccessful.
Since then, going public has been a thorn in Bitmain’s side. Only recently, in 2025, did this largest global manufacturer of crypto mining hardware finally find a new path into the capital markets.
In 2022, Antalpha was founded to provide supply chain financing solutions for the digital asset industry, especially crypto mining. Essentially, supply chain financing involves loans to miners to help purchase mining machines, loans collateralized by mined Bitcoin, and financing based on hash rate—helping miners with cash flow issues.
Initially, Antalpha was under the same parent company as its sister company Northstar, both owned by Jihan Zhang. After restructuring in 2024, it operated independently as “Bitmain’s core financing partner,” and on May 6 last year, it listed on Nasdaq with an IPO price of $12.8. On the first day, Antalpha’s stock soared to nearly $28, with a market cap of $660 million, but now its stock price has fallen to around $9.3.
Just a month before Antalpha’s IPO, Cangu agreed to sell its automotive-related business in China for $352 million to Ursalpha Digital Limited, a buyer introduced by another company, Enduring Wealth Capital Limited (EWCL).
These two seemingly unrelated companies are actually connected through multiple links.
According to TheMinerMag, Hong Kong company records show that Ursalpha Digital, which took over Cangu’s automotive business, shares the same address as Antalpha Digital Limited. Its director, Chuan-Wei Kiu, founder of Taiwan Wanchai Securities, is also a director of Antalpha Singapore.
The “kind introducer,” EWCL, remains somewhat mysterious, registered in the British Virgin Islands, so its shareholder info is unavailable. However, a document submitted by Cangu to the SEC last March revealed that EWCL’s directors are Andrea Dal Mas, Peng Yu, and Anggun Mulia Fortunata. Andrea Dal Mas and Anggun Mulia Fortunata are two of the three actual controllers of Antalpha.
This document states that EWCL plans to acquire 10 million Class B common shares from Zhang Xiaojun and Lin Jiayuan of Cangu for cash. Each Class B share has 20 votes, meaning EWCL—or perhaps Antalpha or even Bitmain—is aiming to acquire control of Cangu.
So, we can sketch a rough picture: after Bitmain’s founders split, Jihan Zhang launched Antalpha in 2022, then pushed for its independence in 2024, and last year used its “independence” to have a related company (Ursalpha Digital) acquire Cangu’s domestic auto business, and another related company (EWCL) to acquire control of Cangu.
This is indeed clever capital maneuvering, but the story is far from over. Let’s continue.
0.39% Restraint
If Bitmain’s failure in Hong Kong stocks was due to internal chaos, then its rumored failure in the US market might be because it accidentally entered the storm of great power rivalry.
The core technology of Bitcoin miners comes from chips. As early as Trump’s predecessor’s term, the US government harbored deep hostility toward a Chinese tech company hoping to list in the US. Although Bitcoin mining chips simply embed mining algorithms into hardware, they still trigger reactions from some US politicians. Over the past few years, rumors of backdoors in Bitmain’s Antminer have often come from North America, with obvious intent.
Therefore, if Bitmain indeed lost the US listing opportunity due to this reason, and if it hoped to at least go public via a shell of Cangu, a prerequisite is that Cangu can no longer be a Chinese concept stock.
Three weeks after Antalpha’s listing, Cangu announced it had completed a board personnel adjustment after selling its auto business and would apply to the China Securities Regulatory Commission to cancel its “Chinese concept stock” status. Cangu stated that its core business had shifted to Bitcoin mining, with operations covering North America, the Middle East, South America, and East Africa. On October 15, Cangu decided to terminate its plan to list via depositary receipts and directly list Class A shares on NYSE. Less than a month later, on November 6, Cangu officially announced it would provide Class A share trading on NYSE starting November 17, and is also actively seeking to enter high-performance AI computing.
What might have prompted all this is perhaps an early plan, or it could be due to the call from US Republican Congressman Zachary Nunn in early September to investigate Bitmain and Cangu. Nunn said that Bitmain and Cangu “seem to be expanding their US operations through complex ownership structures and financing arrangements, which regulators and the public may not fully understand.” Both companies’ representatives responded that they strictly comply with US laws and have no connection with any government or state-owned enterprise.
See? We are not the only ones aware of this.
Even if compliance issues seem to have been addressed, Bitmain has not acted rashly. On December 29, Cangu announced that EWCL would subscribe to 7 million Class B shares at $1.05 million. After the transaction, EWCL’s stake in Cangu would increase from about 2.81% to 4.69%, with voting rights rising from approximately 36.68% to 49.61%. The deal is expected to close in January this year.
49.61%, just 0.39% away from controlling the company. That 0.39% might be Bitmain’s way of testing US regulatory attitudes.
At this point, all clues are connected. Hardware products like mining machines containing chips may prevent sales in the US capital market in the foreseeable future, but mining operations still have a chance. These interconnected maneuvers, if not through a shell listing, can only be called a clever coincidence.
In fact, the first signs appeared as early as last March. In that month, Cangu mined 530 Bitcoin, which, based on total network hash rate, would require at least 29 EH/s. By November 2024, after paying for 32 EH/s of hash rate, nearly all of it was operational by March 2025. This indicates that the mining machines purchased from Bitmain, worth over $250 million, are already in operation, not just bought and stored, but already set up and running.
The Last Lifeline: Providing Computing Power for AI
Bitmain’s second-largest competitor, Canaan Creative, has a total market value of around $550 million. Even if Bitmain successfully lists, it is unlikely to return to its 2018 peak of $15 billion. But if it leverages its mining farms and power resources to provide computing power for AI giants, the story could be different.
MSX analyst Daisy pointed out that if Bitmain only lists as a pure mining farm, the result won’t be very good; it would have to transform and sell its existing power and computing resources. This story needs real orders to be credible. Daisy cited Mara as a typical example: although it aims to pivot to AI computing services, it has no actual orders, so no one is buying. Companies with power resources and real orders, like Applied Digital and IREN Ltd, perform much better, with market caps of about $9 billion and $15.45 billion, respectively.
If Cangu can secure real orders, perhaps Bitmain can leverage the AI arms race to give its former investors a long-overdue explanation after eight years. Currently, Cangu’s market value is about $500 million, and this 30-fold climb still requires gradual effort.
