Canaan's 1,117 Pledged Bitcoin Are Not a Treasury: What the $97.6 Million Quarterly Loss Actually Says

Consider the moment when a Bitcoin miner discovers that its treasure chest is also a lien. That moment arrived on Sept. 8, when Canaan — one of the oldest makers of Bitcoin mining machines — reported $31.9 million in second-quarter revenue, missing the $35 million to $45 million range it had promised in May. The shortfall was not subtle. Product revenue collapsed to $13.6 million from $42.9 million in the first quarter, a 68% drop that the company blamed on weaker demand, lower computing power sold, and falling selling prices. Management then guided third-quarter revenue to just $11 million to $15 million. In a bull market that has taught investors to celebrate hashrate, machine sales, and digital-asset treasuries, Canaan's numbers read like a reminder that mining hardware is a cyclical industrial business before it is a crypto story. But the headline revenue miss is not the most consequential number in the release. The more important detail hides in the treasury footnote: as of June 30, Canaan held 1,915.5 Bitcoin, yet more than half of those coins — 1,117 BTC to be precise — were pledged as collateral for secured term loans. Another 100 BTC had been transferred into a fixed-term product. Only 698.5 BTC sat in the company's unrestricted cryptocurrency assets category. The boldest line in the report is not the net loss of $97.6 million. It is the quiet admission that a Bitcoin balance is not a pool of spending power when a lender holds the key. I have spent years auditing the economic models of failed crypto projects, largely during the painful unwind of 2022. In that work, I developed a habit that has never failed me: I ignore the headline treasury number and go straight to the collateral rows. The distinction between "we own Bitcoin" and "we control Bitcoin" is where mining companies hide their true risk. Canaan's latest filing is a textbook case. The context matters because Canaan is not a marginal player. Founded in 2013, the company survived multiple crypto winters and became a public benchmark for the ASIC manufacturing industry. It designs and sells application-specific integrated circuits that power Bitcoin mining farms, and in recent years it pivoted aggressively into self-mining — running its own machines rather than merely selling them to others. That pivot looked brilliant when Bitcoin climbed and the company accumulated a record digital-asset treasury. Related reporting from earlier this year put Canaan's combined BTC and ETH holdings near $148 million, a figure that generated considerable enthusiasm among retail observers who track miner balance sheets as a proxy for conviction. The enthusiasm always missed a crucial layer. A miner's corporate treasury is not a personal cold wallet. It is an operating account, a collateral base, a buffer for inventory cycles, and occasionally a source of last-resort liquidity. When hardware sales slow, the treasury becomes the business. When the treasury itself is encumbered, the distinction between what a company owns and what it can actually spend becomes the entire story. Let me break down the balance sheet with the precision that the second-quarter report demands. Canaan's total Bitcoin holdings at June 30 were 1,915.5 BTC. Of those, 1,117 BTC were pledged as collateral for secured term loans. That is 58.3% of the entire Bitcoin position. An additional 100 BTC sat inside a fixed-term product, meaning they were committed to a contractual arrangement with a maturity constraint. So 1,217 BTC, or 63.5% of the reported holdings, were not freely available in the way that most investors imagine when they read "treasury." The remaining 698.5 BTC constituted the company's cryptocurrency assets classification. Accountants built a wall between these categories for a reason. The 1,117 pledged and 100 fixed-term coins were recorded as cryptocurrency receivables valued at $70.9 million at June 30. The 698.5 unrestricted coins were recorded separately as cryptocurrency assets valued at $47 million. Those two lines tell completely different stories. The $70.9 million receivable figure is not a pot of money waiting to be spent. It is an accounting acknowledgment that the coins have been leveraged, promised, or otherwise committed. In a liquidation scenario, those coins do not belong to Canaan first; they belong to the creditor first. This is the kind of nuance that bull markets routinely erase. When Bitcoin is rising, a miner that announces a growing BTC balance is celebrated as a diamond-handed accumulator. The same announcement, read through the lens of collateral, may actually describe a company that is increasingly borrowing against its own future. Pledging Bitcoin as loan collateral is not inherently reckless — in a rising price environment, it can be a rational way to access capital without selling coins. But it transforms the nature of the treasury from a store of value into a margin account. The moment Bitcoin price drops, pledged coins trigger maintenance calls. The miner must then either supply additional coins or watch the lender liquidate the position at the worst possible moment. That dynamic is not hypothetical; it is the precise mechanism that destroyed several over-leveraged mining operations during previous downturns. What makes Canaan's position notable is the sheer proportion. When 58.3% of your Bitcoin is already serving as loan collateral, a sharp correction does not merely reduce your net worth — it forces decisions about which obligations to prioritize. The fixed-term product adds another layer of rigidity. Coins locked in fixed-term arrangements cannot be deployed to meet margin calls, pay suppliers, or fund operations without incurring penalties or breaching contractual terms. The market, however, tends to look at one number. I saw this pattern repeatedly during the 2022 collapse, when projects with massive treasuries and immaculate narratives were actually weeks away from insolvency. The difference between a treasury and a trap is almost always found in the footnotes that nobody wants to read. The quarterly loss itself demands careful unpacking, because it looks far worse — and in some ways far better — than the cash dynamics suggest. Canaan reported a net loss of $97.6 million for the second quarter. That is a staggering figure for a company whose revenue was only $31.9 million. But a net loss computed under accounting rules is not the same as cash burned. The loss included $25.3 million in inventory and prepayment write-downs plus purchase-commitment provisions. It also included $9.2 million in property and equipment impairment. Together, these disclosed noncash charges totaled $34.5 million. The remaining loss components likely included depreciation, fair-value changes on digital assets, and other noncash items that do not directly deplete the bank account. Here is the surprising twist: Canaan's cash actually increased during the quarter. The company reported $66 million in cash at June 30, up from $43.5 million at March 31. That increase came despite the massive accounting loss, which tells us that cash inflows from loans, asset sales, or other financing activities outweighed operational cash outflows during those three months. But context is crucial: $66 million is still comfortably below the $80.8 million the company held at the end of 2025. The trajectory is not a straight line upward; it is a wavy line with a downward bias. Cash rising while an accounting loss widens is one of the more misleading optics in corporate reporting. It invites the question: if the company lost $97.6 million, why does it have more cash? The answer is that the loss is paper, the write-downs are real but noncash, and the cash is partly borrowed. Investors who focus solely on the cash balance without examining the liabilities side are repeating the same mistake they made with over-leveraged miners in previous cycles. Mining operations still contributed meaningfully to the top line. Canaan produced 243 BTC during the quarter and generated $17.7 million in mining revenue. Management emphasized that these operations made a positive cash contribution before depreciation — a carefully chosen framing that excludes the very real cost of replacing worn-out machines and maintaining infrastructure. Positive cash contribution before depreciation is a much narrower measure than true profitability, because depreciation exists for a reason. Mining rigs do not last forever. Every ASIC has a finite useful life, and a company that cannot generate enough cash to replace its machines is slowly liquidating its productive capacity. The Ethiopia situation illustrates this concern: related reporting noted that Canaan's paused Ethiopia mining operation accounted for nearly 35% of its July operating hashrate total. When a third of your operating hashrate is sitting idle, the distinction between installed capacity and productive capacity becomes brutally visible. The product business, meanwhile, is contracting faster than the mining business can compensate. Second-quarter product revenue of $13.6 million was less than one-third of the first-quarter level. Lower computing power sold means fewer new customers; lower selling prices mean the competitive position is eroding; and the $25.3 million write-down suggests the company is holding inventory that it cannot sell at the price it once expected. In the ASIC industry, inventory write-downs are a leading indicator of technological obsolescence. If Canaan cannot sell its current generation of machines, it either has too much supply or its machines are no longer competitive against newer, more efficient models. Either explanation points to a structural problem, not a temporary hiccup. Management's third-quarter guidance makes the trajectory unmistakable. Revenue of $11 million to $15 million implies another sequential decline of 53% to 66% from the second quarter. No amount of narrative framing can dress up a 66% revenue contraction. The question is no longer whether Canaan is facing a cyclical downturn; it is whether the company's balance sheet can absorb the downturn without forcing the liquidation of its remaining assets. That brings us to the most revealing event of all: the decision, disclosed in September, to sell digital assets after the quarter ended. Canaan said it sold 3,952 ETH and 54 BTC in late August for approximately $13.9 million. Part of the proceeds was used for share repurchases. By Sept. 8, the company's buyback program had repurchased about 16.4 million American depositary shares for $7.4 million, including $5.4 million spent in late August. Read those numbers carefully. The company sold Ether and Bitcoin to fund share buybacks. In a vacuum, buybacks are a legitimate capital-allocation tool. When management believes its stock is undervalued, repurchasing shares can return value to remaining shareholders and signal confidence. Related reporting even suggested that Canaan's crypto holdings were large enough to buy back nearly 20% of its market value while its core business burned cash. That observation sounds bullish on the surface: the company has a war chest, and it is deploying that war chest to support its equity. But the timing exposes a contradiction. Canaan missed its revenue guidance, guided down sharply, wrote down millions in inventory, pledged more than half its Bitcoin as collateral, and then sold digital assets to buy stock. The market read this as a company defending its share price. A more skeptical reading is that a company in operational distress is choosing to support its equity rather than strengthen its balance sheet. In a quarter when suppliers are demanding cash, when inventory is being written down, and when mining operations are partially paused, spending $7.4 million on repurchases is a statement about priorities. The contrarian angle is worth stating plainly: buybacks are not automatically reckless. If Canaan's market capitalization fell below the value of its net cash plus its unencumbered crypto assets, repurchasing stock could be mathematically rational — a way to retire equity at a discount to intrinsic asset value. In that scenario, selling $13.9 million of crypto to capture a larger ownership stake at depressed prices might be the most disciplined allocation management could make. The shareholders who remain after the buyback own a larger slice of a company that still holds hundreds of Bitcoin, millions in cash, and a functioning — if challenged — mining operation. Yet the value-first analysis cuts against that rationale. A company with $66 million in cash, a $97.6 million quarterly loss, shrinking product revenue, and a third-quarter guide that implies continued contraction is not in a position to prioritize share price over structural survival. Every dollar spent on buybacks is a dollar not spent on developing next-generation mining chips, not spent on restarting operations in Ethiopia, and not spent on reducing the collateralized debt that encumbers more than half the Bitcoin treasury. In previous downturns, the companies that survived were not the ones that defended their stock prices; they were the ones that hoarded liquidity, cut costs early, and avoided the appearance of financial strength while their operations deteriorated underneath. The deeper issue is linguistic. We call Canaan's Bitcoin holdings a "treasury," but a treasury is supposed to be a reserve of last resort — assets that can be deployed quickly in a crisis. A Bitcoin that is pledged as collateral is not a reserve; it is a contractual obligation with a variable margin requirement. A Bitcoin locked in a fixed-term product is not liquid; it is committed. Only the 698.5 BTC sitting in unrestricted cryptocurrency assets function as a true reserve, and even those coins are subject to price volatility that can erode their value overnight. This is why I have grown uncomfortable with the industry's habit of celebrating miner treasuries during bull markets. The celebrations ignore the difference between gross holdings and net unencumbered holdings. They ignore the counterparty risk embedded in every loan agreement. They ignore the possibility that a miner's largest asset is actually its largest liability waiting to be triggered. Bull markets reward confidence, but they punish the failure to distinguish between ownership and control. Canaan's own accounting already makes that distinction. The company separated cryptocurrency receivables from cryptocurrency assets, and it separated both from cash. It provided enough information for a careful reader to understand that 1,915.5 BTC does not mean 1,915.5 BTC of freedom. The problem is that the market's vocabulary has not caught up with the accounting. Every headline that screams "miner adds Bitcoin to treasury" is an invitation to ignore the pledged coins sitting right below the headline. Based on my audit experience during the last bear market, the first thing I check in any mining company's balance sheet is the encumbrance ratio: the proportion of digital assets that are pledged, loaned, or otherwise restricted. Canaan's ratio is 58.3% as of June 30, and that is before considering the additional 100 BTC locked in the fixed-term product. The unencumbered ratio — Bitcoin that could theoretically be sold or deployed without asking a lender's permission — is just 36.5%. A company whose entire strategic narrative is built on Bitcoin conviction is, in reality, a company that has borrowed against more than half of its Bitcoin position. That is not conviction; it is leverage wearing a conviction costume. The third-quarter guidance of $11 million to $15 million creates a stark arithmetic problem. If product revenue continues to shrink, Canaan will lean more heavily on mining revenue. But mining revenue depends on Bitcoin price and operating hashrate, and July data shows that a significant portion of the company's hashrate was paused in Ethiopia. Management says mining still generates positive cash contribution before depreciation, but that narrow measure does not capture the full cost structure. When the depreciation of hardware is included, when inventory write-downs are included, and when the interest costs on secured loans are included, the picture darkens considerably. The late-August crypto sale does offer one piece of genuine comfort: Canaan demonstrated that it could convert digital assets into cash quickly when needed. The $13.9 million raised from selling 3,952 ETH and 54 BTC shows that crypto markets remain liquid enough to absorb miner sales without catastrophic price impact. In a crisis, that liquidity is valuable. But it also reveals a fundamental shift in how Canaan views its digital assets. Once a company begins selling Ethereum to fund share buybacks, its Bitcoin holdings are no longer a strategic reserve; they are a funding source. The psychology has changed, even if the treasury headline has not. There is a painful irony here. Canaan's pivot into self-mining was supposed to reduce its dependence on hardware sales and give it direct exposure to Bitcoin's upside. Instead, the pivot created a company that needs Bitcoin to stay high, needs machine sales to recover, needs loan covenants to remain loose, and needs its share price to justify continued buybacks. That is not a robust business model; it is a bundle of correlated bets. When Bitcoin falls, pledged coins trigger margin calls. When Bitcoin falls, mining revenue falls. When Bitcoin falls, the value of the cryptocurrency assets on the balance sheet falls. When Bitcoin falls, the equity cushion that supports the buyback thesis falls. Every one of Canaan's strategies is tied to the same asset price, which means the company has not diversified its risk at all — it has concentrated it. The most important lesson from this earnings report is not about Canaan specifically; it is about how the entire mining sector communicates financial health during a bull market. The industry has trained investors to focus on hashrate growth, production volume, and treasury accumulation. Those metrics matter, but they are incomplete. Hashrate does not capture the cost of debt. Production volume does not capture the price at which machines were bought. Treasury accumulation does not capture the percentage of coins that are pledged to lenders. The complete picture requires all of those layers together. A pledge is not a promise. A treasury is not a revenue line. And a coin that belongs to a creditor in a downside scenario is not a coin that belongs to the company. Canaan's accountants understood this when they divided the balance sheet into receivables and assets. The company's management understood this when they sold 3,952 ETH to raise cash. The only question is whether the broader market will understand it before the next major correction arrives. The loss of $97.6 million is a headline number, but headlines do not capture the texture of a company in transition. They do not reveal that cash rose to $66 million even while the loss widened. They do not explain that the increase came from financing activities rather than sustainable operations. They do not show that the third-quarter guidance implies another crushing sequential decline. To understand Canaan, one must ignore the headlines and read the footnotes — the same way one must ignore the total Bitcoin balance and count the pledged coins. I have watched enough mining companies fail to know that the ones that survive are rarely the ones with the largest treasuries. They are the ones with the smallest encumbrance ratios, the clearest cost structures, and the greatest willingness to sell assets early when the market still allows them to do so at reasonable prices. Canaan has now shown that it can sell crypto when necessary. The test ahead is whether it can manage its remaining treasury, its debt obligations, and its shrinking product business without being forced to sell at the worst possible moment. Bull markets forgive leverage; bear markets punish it without mercy. The companies that remember this during the euphoric phases are the ones that remain standing when the cycle turns. Canaan's second-quarter report is a reminder that even in a bull market, some balance sheets are closer to the edge than their Bitcoin balances suggest. Count the unencumbered coins next time. Count the receivables. Count the pledged collateral before you count the conviction. Because in the end, the only treasury that matters is the one you can actually use when everything else fails. By that measure, Canaan's real treasury is not 1,915.5 Bitcoin. It is 698.5 Bitcoin, $66 million in cash, and an uncertain path back to profitability in a market that no longer seems eager to buy what the company is selling.