Canaan Sells Its Bitcoin to Buy Itself: A Capital Allocator's Confession

I used to think mining companies were the most unshakeable Bitcoin believers. They run the machines. They pay the power bills. They watch the hashrate like a heartbeat. So when Canaan — the company that manufactures the Avalon ASIC miners — announced it would sell crypto assets to fund a $30 million share repurchase, the headline felt like a footnote. A routine corporate finance disclosure on a Nasdaq newswire, the kind of thing you scroll past. But the footnote is the confession. Canaan's board approved a plan to liquidate part of its cryptocurrency holdings to finance the repurchase of its own Nasdaq-traded shares (ticker: CAN). The company did not disclose the exact composition of its digital asset treasury. As a miner and hardware vendor, though, it is almost certain to hold Bitcoin — receivables from customers who pay in BTC, rewards from its own mining fleet, and legacy positions acquired during earlier cycles. The announcement is not a technical upgrade. There is no new chip. No protocol change. No network launch. It is a corporate finance decision with a quiet power: it tells you what management thinks about its own stock versus the asset its products are designed to dig out of the earth. To understand why this matters, you have to remember where Canaan sits. Alongside Bitmain and MicroBT, it is one of the three surviving ASIC giants — a small club that has outlasted every other manufacturer of SHA-256 mining hardware. Bitmain is private, with a notoriously complex internal structure. MicroBT is private as well. Canaan is the only one playing in the sunlight of a Nasdaq listing, which makes its capital allocation decisions more visible and, in some ways, more consequential than its competitors' private maneuvers. When a company in this club rebalances its treasury, the rest of the mining industry reads the tea leaves. I spent 2017 manually reviewing Solidity code while others chased ICO flips. What I learned then was that decentralization requires rigorous engineering, not just good intentions. The same test applies to balance sheets. When a miner sells the thing it mines to buy a claim on its own future, you have to audit the governance, the timing, and the fear behind the move. Here is the core arithmetic of Canaan's swap. First, look at the supply side. A share repurchase reduces the float. If Canaan retires $30 million of its shares, every remaining share holds a slightly larger claim on future earnings. That is the textbook argument, and it is not dishonest. Buybacks historically offer moderate price support in the weeks after announcement — not because the market suddenly believes in the business, but because a smaller supply of shares meets an unchanged demand curve. The signal is real: management is telling the market it believes CAN is undervalued relative to its own estimates of cash flows from hardware sales and mining. Second, look at the sell side. Canaan is disposing of Bitcoin. This is the part most crypto-native readers will skim past, but it is the part that matters most. Under U.S. GAAP, crypto assets held by public companies were, for years, treated as indefinite-lived intangible assets. That meant no mark-to-market gains. A company could buy Bitcoin at $20,000, watch it trade at $80,000, and still record it at historical cost, subject only to impairment charges if the price fell. The rule created a strange and quiet pressure: the only way to convert an unrealized gain into a visible profit on a public income statement was to sell. I remember sitting with a CFO in early 2021, in the middle of a bull run. He leaned across the table and said, "we can't show the upside, so why are we holding it?" That is the hidden thesis behind so many miner sales. It is not a bearish view on Bitcoin. It is a bearish view on what unrecorded upside does to an audited earnings release and to the institutional investors who are professionally incapable of valuing a balance sheet item they cannot see in the footnotes. The accounting framework shifted slightly when the Financial Accounting Standards Board issued new rules allowing fair value measurement for certain crypto assets, effective for fiscal years beginning after December 15, 2024. Under that regime, companies can mark their holdings to market each quarter. But the older mindset still governs many corporate treasuries, and more importantly, the decision to sell is rarely driven purely by accounting. It is driven by cash need, by risk appetite, and by the board's aggregate judgment about the next twelve months. Third, look at the message. When a public company chooses to buy its own equity with its Bitcoin, it is making an explicit statement: at this margin, on this date, CAN stock offers a better risk-adjusted return than BTC. That is a sentence no pure mining company wants to say out loud, because it undermines the entire "Bitcoin is the ultimate treasury asset" narrative. Marathon Digital and Riot Platforms have spent the past two years going in the opposite direction — holding their treasury coins, raising equity or convertible debt, and in some cases buying more Bitcoin. MicroStrategy represents the extreme version of that conviction, issuing billion-dollar debt tranches to accumulate Bitcoin at scale. Canaan has now positioned itself as the outlier on the other side. Among public miners and mining hardware companies, it is choosing to be a cash-first hardware business, not a Bitcoin proxy. That distinction matters more than the market seems to realize. A dual valuation — mining hardware business plus Bitcoin treasury — is what allowed CAN to trade with Bitcoin's beta. If you held CAN during a bull run, you got leveraged exposure to BTC through a company that also had real earnings. Once the treasury is harvested and converted into a smaller share count, that lever disappears. Canaan becomes a pure semiconductor play on the ASIC cycle. Its earnings per share might improve. Its correlation to Bitcoin will drop. That is a more disciplined trade and a less magical one. I want to pause on the two-token structure, because I think it is the most misunderstood part of the announcement. Canaan has, in effect, two liabilities and two assets on opposite sides of the same table. On one side, a public equity that promises governance rights and residual cash flows. On the other, a private reserve of Bitcoin that promises nothing statutory but offers optionality. A treasury sale converts the second promise into the first. That is an act of translation, not creation. The board is saying: we earned more confidence from the market by buying our own equity than by holding digital gold. In a world where MicroStrategy borrows money to buy Bitcoin, Canaan is the mirror image — it sells Bitcoin to buy itself. Now, the pragmatism test. The contrarian angle. A $30 million buyback is not a large amount for a company that has manufactured and sold thousands of ASIC miners. In the context of Canaan's market capitalization — which over the past four years has swung between a few hundred million and several billion dollars — $30 million is real but not transformative. So why go through the trouble of selling crypto to finance it? Why not simply declare a dividend, or allow the treasury to sit? The answer is in the signal, not the size. The board is speaking to a specific audience: shareholders who have watched the stock bleed out during the bear market cycles. The buyback is a confidence gesture. "We see the value." And that is exactly why it deserves suspicion. Because if Canaan truly believed Bitcoin's upward trajectory was intact, selling any portion of a finite digital stockpile — even to fund healthy buybacks — would be an opportunity cost measured in future gains. The board's decision implies a risk model in which Bitcoin's volatility is not worth the carry. Fine. But it may also imply something less flattering: that the company needs the cash. That is the fear I would follow. Notice the announcement does not say the crypto sale is a discretionary portfolio rebalance. It says the proceeds will fund the repurchase. That construction turns Bitcoin into a funding source, not an asset class. There is a meaningful difference between "we are trimming our Bitcoin position to capture GAAP gains" and "we need liquid reserves to execute a capital return program." Both are legitimate. They produce different headlines and different long-term consequences. The crypto-native reaction will be a mixture of disappointment and relief. Disappointment, because every Bitcoin believer wants a miner to hold forever. Relief, because the sum is small and the company's hardware business is intact. Both reactions miss the deeper point. The announcement is not a trade idea. It is a definition of what kind of company Canaan wants to be. And in a bull market where every second project claims to be "Bitcoin-aligned," a company that quietly sells its treasury to improve its public equity deserves a sober read, not a moral judgment. Canaan's underlying business has faced the same pressures as every hardware vendor in the mining supply chain: order cyclicality, dependence on advanced wafer foundries, and the long shadow of Chinese regulatory policy. Mainland China has effectively banned crypto trading since 2021. The operating entities of Canaan sit inside that jurisdiction. Holding a substantial Bitcoin treasury on a Chinese-linked balance sheet is a compliance headache — repatriating proceeds, justifying holdings to auditors, navigating two legal systems with conflicting signals. Selling reduces friction. A dollar is clean. A Bitcoin, from a Beijing audit committee's perspective, is a question mark. There is also the self-mining duality. Canaan is not only a hardware vendor; it operates its own mining rigs. That makes it, at the same time, a supplier and a competitor to its own customers. When the company sells Bitcoin to fund buybacks, it is also signaling that its self-mining expansion plans are not the top priority right now. Resources are flowing back toward the core hardware business and toward the public market's preference for visible returns. That is a strategic statement, and it is worth reading closely. What does this mean for the wider market? Small but symbolic. $30 million is dust compared to the tens of billions of dollars that change hands in daily Bitcoin spot volume. The sell pressure is negligible. But the narrative pressure is not. Every time a publicly traded miner or mining hardware maker decarbonizes its balance sheet, the marginal buyer of Bitcoin becomes a little more retail, a little more foreign, a little less institutionally anchored. This is how "institutions are accumulating" stories get quietly walked back — not in a single cascade, but in a thousand small treasury decisions that no headline ever covers. There is also a governance question that deserves more ink than it will receive. Who approved the exact timing of the sale? Was the liquidation subject to a standing treasury policy, or was it a discretionary board action? For a company founded in Beijing and listed in New York, the answer matters. SEC rules like Rule 10b-18 create a safe harbor for share repurchases, but they do not deeply scrutinize the source of funds. The structure here — selling one asset to buy a claim on the same company — resembles a capital allocation trade. Capital allocation trades deserve the same scrutiny as smart contract audits. The code is a promise; the board's minutes are a promise with a lower timeout. I keep returning to my own history with this industry. In 2020, during DeFi Summer, I watched the mechanisms of incentive design turn into vehicles for pain. Compound's governance token crash wiped out savings that people in my Beijing study group had worked a year to accumulate. The lesson I absorbed was not that decentralized finance is worthless. It is that yield can be mathematically elegant and morally destructive at the same time. When I look at Canaan's treasury move, I see the same pattern in miniature: an elegant financial mechanism — sell crypto, buy back stock — that converts a long-term belief into a short-term favor. The choice is not wrong. But it is a choice, and it deserves to be called what it is. Here is what I will be watching in the coming quarters. Not the next CAN earnings report, though I will read it. Not the daily BTC chart, because in a bull market the chart always flatters. I will be watching the next public miner's treasury statement. If another miner follows Canaan and sells Bitcoin to fund a buyback, the story is not "shareholder returns." It is sector-wide deleveraging of crypto exposure by the very companies claiming to mine because they believe. If instead the peers hold and accumulate, then Canaan's sale will look like a one-off hedge — a rational treasurer's move in an irrational market. I do not pretend to know whether CAN is a better buy than BTC at this moment. My rule has always been to follow the fear, not the chart, and my fear here is that the market will file this under "noise" and move on. But a trend that starts as noise ends as structure. Watch the second miner. Watch the third. If the treasury-sale playbook spreads, the public mining sector will finish this cycle with far fewer coins on its balance sheets than the narrative suggests — and the "Bitcoin treasury" theme will migrate to a shrinking group of pure holders. Follow the fear, not the chart. The fear in Canaan's boardroom was not about the share price. It was about a future in which Bitcoin's unrealized gains could not be recorded on a statement, and in which a shareholder base conditioned by two brutal bear markets demands visible earnings now. That fear is real. It is also, in my experience, the most common reason good people execute mediocre capital decisions. The question for the rest of us is simpler. If you can measure the conviction of a miner by what it does with its product, then measure Canaan not by the terahashes its machines promise, nor by the dust of a $30 million repurchase, but by the quiet admission hidden in this announcement. The company that makes the shovels now wants to own a smaller part of the mine. That is a sentence worth rereading when the next bull market peaks and every miner is buying jets.