When a Miner Sells Its Own Hoard: The Canaan Buyback Confession

The SEC form hit EDGAR in the quiet hours of an August trading day. 2:55 p.m. EDT, August 4. Canaan Inc. — the second-oldest Bitcoin mining hardware maker on the Nasdaq — announced that management could now sell part of its roughly $130 million digital asset treasury to fund an existing share buyback program. Most coverage treated it as corporate finance routine: a struggling company unpacks a little liquidity. Nobody seemed to notice that the filing was, in its own quiet way, a confession.

At that same moment, Canaan's market capitalization sat at $144.7 million. Now add the pieces: roughly $130 million in BTC and ETH at August 3 prices, plus $43.5 million of cash reported on March 31. The gross sum of those inputs is $173.5 million. Do the subtraction. The entire public company was trading roughly $28.8 million, or about 19.9%, below the value of its own treasury plus cash. A nearly twenty percent discount on a business that owns a working ASIC manufacturing operation, a facility build-out, and a mining fleet. That discount wasn't a gift to investors. It was a warning light.

Because the way Canaan chose to address that discount wasn't through operational excellence. It wasn't through a hardware breakthrough. It was through permission to sell the company's Bitcoin and Ethereum into the market to buy back its own shares. Minted in hope, burned in regret. But this time, the regret might be double-edged: not only the coins sold, but the company that sold them.

The code didn't change. Business metrics didn't change. A single paragraph of disclosure changed the entire risk profile of a company that already lost $88.7 million in one quarter. That's the kind of moment I built my career on — where the public narrative and the underlying ledger disagree. And they are disagreeing loudly.

CONTEXT

Let me lay out the stage before I start cutting. Canaan makes ASICs — the application-specific integrated circuits that power Bitcoin mining. Founded in Hangzhou, the company has shipped millions of units across its Avalon series. It listed in New York in 2019 via one of the earliest Chinese crypto-adjacent IPOs. For most of its history, Canaan's stock has been a leveraged proxy on Bitcoin's price with a hardware-sales multiplier attached. When BTC rallies, miners order more rigs, and Canaan prints revenue. When BTC stagnates, orders shrink, ASPs compress, and the company bleeds.

2026 was supposed to be the year the cycle turned. It didn't. The first quarter of 2026 delivered a $22.9 million gross loss, a $54.3 million operating loss, and a net loss of $88.7 million. Second-quarter revenue guidance of $35 million to $45 million was in line with expectations, but "in line" in a bear cycle means the bleeding has slowed, not stopped. The hardware segment — the core engine of the business — is facing collapsed demand for ASIC rigs. Miners are turning off unprofitable machines. Data-center operators are delaying fleet expansions. And every unsold Avalon machine in a warehouse is a piece of metal with rapidly declining economic value.

But here is the twist that makes Canaan interesting to me, and not just another distressed micro-cap: the balance sheet. As of June 30, Canaan held 1,915 BTC and 3,952 ETH, including receivables and excluding customer deposits. At August 3 prices, that block of digital assets was worth approximately $130 million. The cobalt reality is that the company's digital treasury is now worth more than many parts of its actual business. In better cycles, Canaan's mining rigs were its crown jewels. In this cycle, the crown jewel is the Bitcoin accumulation of its treasury desk.

That transformation is happening across the industry. CoinShares' latest mining report flags the shift: stressed miners are selling coins, stronger operators are pivoting into AI, and listed mining stocks are becoming less pure Bitcoin proxies than many investors assume. Canaan sits in neither clean bucket. It is not strong enough to pivot cleanly into AI infrastructure without diluting shareholders. It is not wealthy enough to HODL through the cycle like Marathon or Riot with their fortress balance sheets. Instead, Canaan has chosen a third path: an internal capital allocation loop where BTC and ETH fund the repurchase of Canaan equity.

The Aug. 4 SEC filing is the mechanism. The Dec. 12, 2025 board authorization is the container. By May 19, Canaan had spent about $2 million repurchasing roughly 2.8 million ADSs under the $30 million, twelve-month program. Simple subtraction puts the remaining nominal capacity at approximately $28 million as of May 19. The Aug. 4 filing did not disclose newer repurchase totals, so the live unused authorization remains a moving variable. The program runs through the end of 2026 — roughly four more months of runway at the time of writing. And the company just made the funding source explicit: the digital asset treasury, unpledged or at least partially unpledged, can be sold to pay for share buybacks.

I have to pause here. Because I've audited protocols and balance sheets in enough cycles to know that the words "treasury sale to fund buybacks" are never neutral. They mean one of two things. Either management believes its own shares are dramatically undervalued and is deploying real assets at a discount — a rational trade. Or management needs a way to support the stock price without raising new capital, because the equity market won't lend it money at reasonable rates, and the debt market has already extracted the collateral it can.

Both explanations deserve a forensic read.

THE GEOMETRY OF THE FILING'S SILENCES

The first thing a cold dissection of the Aug. 4 filing surfaces is what it doesn't say. No amount of treasury assets to be sold. No dollar ceiling on the liquidation beyond the $28 million remaining buyback capacity. No schedule. No disclosure mechanism for when sales occur. No requirement to notify shareholders at the moment BTC is moved to an exchange. Just an authorization, shaped like a rubber stamp, placed inside a 6-K.

In my line of work, disclosure silences are the most valuable data in the dataset. When a company leaves the "how much" and "when" out of an announcement, it is not being careless. It is reserving optionality. Canaan's management team gains the ability to sell exactly as much crypto as the buyback demands, at times of their choosing, with no real-time transparency obligations. Shareholders only learn of the conversion when the next quarterly report lands — weeks or months after the on-chain transaction actually occurred.

This is meaningful because the sequence matters. If Canaan sells large blocks of BTC through liquid venues, it depresses the market. If it executes carefully through OTC desks or block trades, the impact can be minimized. The announcement itself doesn't tell us which route management will choose. But here's what it does tell us: they have decided that their own stock is a better investment than the Bitcoin they've been accumulating since the bear market began. That is a startling statement of conviction, and most analysts glossed over it completely.

Let me put a number on it. The buyback program has a nominal ceiling of $30 million, of which roughly $28 million may remain. Canaan's treasury is approximately $130 million. So the maximum drawdown from the treasury equals roughly 21.5% of the current digital asset book. It is not a full liquidation — not even close. But it filters through the lens of the company's operating burn. At Q1's run-rate, $28 million is less than one quarter of the total losses. It is approximately 28% of the annual expected burn if losses hold near $100 million. The program is not designed to save the company. It is designed to do something subtler: steady the share price and reduce the overhang of available float, in the hope that the market notices the company's assets are worth more than its stock.

That argument has a real mathematical core. Buybacks at a 19.9% discount to gross asset value are mechanically accretive to per-share value — if the assets backing the buyback were not simultaneously being depleted. But in Canaan's case, the asset backing the buyback is the very treasury that makes up a material fraction of the balance sheet. Every dollar spent buying shares removes a dollar from the pool of corporate assets. The numerator shrinks as the denominator shrinks. Whether the per-share value increases depends on the exact rate of shrinkage on both sides. If they buy back $2 of shares at a 20% discount to NAV, and they spend $2 of unpledged BTC to do it, the net effect on per-share NAV is a small accretion. But if they sell the BTC at a discount to its mark-to-market value — through a hurried liquidation or a distressed OTC trade — the accretion can vanish entirely. And if the share price fails to appreciate, the buyback becomes little more than a transfer from the treasury to exiting shareholders. The HODLers get diluted in asset coverage. The sellers get liquidity. Management gets a headline.

This is a shell game with one shell and no pea. The pea is the question of whether Canaan's operating business can ever become cash-flow positive again on its own.

THE TREASURY UNDER THE KALEIDOSCOPE

So let's take the ledger apart. Piece by piece.

The June 30 operating update lists 1,915 BTC and 3,952 ETH. This includes receivables. It excludes customer deposits — those go straight back to customers, not into the corporate treasury. At the market prices prevailing in early August, with the industry's metrics clustered around a total of roughly $127 million to $132 million depending on the exact close, the mix runs roughly three-quarters Bitcoin by value and one-quarter Ethereum. The exact split isn't crucial for this analysis; what's crucial is the taxonomy of what is liquid.

In March, Canaan pledged 905 BTC for secured term loans. It placed another 100 BTC into a fixed-term product. That's 1,005 of the 1,915 BTC either encumbered as collateral or locked up in a time-deposit style instrument. In other words, nearly 53% of the Bitcoin reserve was not freely available for sale at the time of the March disclosure. The remaining 910 BTC is what I would describe as the "free float" of the treasury. Pledged collateral can't be sold without ending a loan facility or replacing it with other assurance. Fixed-term crypto deposits typically carry liquidity penalties or lock-up windows that make emergency liquidation costly. So when the Aug. 4 filing authorizes the sale of treasury assets, it is — in practical terms — placing most of the burden on the unpledged slice of the reserve: roughly 910 BTC and whatever ETH is not similarly constrained.

Ethereum, in this structure, is the more natural liquidation asset. It is less likely to be collateralized if the senior secured lenders specifically wanted BTC-denominated collateral, as the March data suggests. ETH held by Canaan appears to be a pure treasury play: 3,952 tokens. At steady prices, a full liquidation of the ETH book would raise around a quarter of the program's remaining capacity — roughly $7 million against $28 million. The rest would have to come from the 910 unpledged BTC, or the pledge would have to be renegotiated. That's a significant operational constraint. The company's capacity to deploy the full buyback is not a simple function of its $130 million headline treasury; it's a function of its collateral web and outstanding debts.

And that collateral web is tightening. At March 31, Canaan had $43.5 million in cash against $106.4 million in current liabilities. Year-end cash had been $80.8 million; the difference tells you the Q1 burn rate consumed roughly $37.3 million of its cash buffer. The company collects customer receivables — April brought in about $42 million — but those inflows are being absorbed by payables, inventory financing, and operating costs. The working-capital position is the real story here. Canaan is a business with obligations coming due faster than cash is arriving. It has one major asset class liquid enough to settle those obligations without restructuring: the cryptocurrency treasury.

The word "treasury" sounds like wealth. In Canaan's case, it is more accurately described as the company's emergency reserve, pension fund, contingency pool, and war chest rolled into one. When a business with negative working capital dedicates that reserve to share buybacks, it is effectively saying: "We believe the stock, not our operational recovery, is the best deployment of our last dry powder." Either that conviction is deeply informed — and the discount is truly temporary — or it is a rationalization for avoiding the harder choices: halting factory expansions, selling the mining fleet, or finding a strategic partner.

THE BUYBACK GEOMETRY AND THE EXIT RAMP

Let me walk you through the per-share arithmetic because the numbers clarify more than the narrative.

Suppose, for illustration, Canaan has roughly 95 million ADS-equivalent shares outstanding. I'm triangulating from the 2.8 million ADSs repurchased and the market cap data; the exact count isn't disclosed in the filing, and my point holds directionally regardless of the precise figure. At a $144.7 million market cap, the price per ADS is around $1.50. The company's gross assets — treasury plus cash — come to $173.5 million. Per share, that's approximately $1.83 of gross asset coverage. The market is pricing in a haircut of roughly 18 cents per dollar.

Now the board spends $2 million buying back shares. At an average price of, say, $1.50, that retires about 1.33 million ADSs. The remaining share count drops. Per-share gross asset coverage rises — if the assets themselves don't change. But the $2 million spent came from the treasury. Treasury value drops from $130 million to $128 million. Cash stays at $43.5 million. Total gross assets: $171.5 million. Divided by fewer shares, per-share coverage works out slightly better than before. Good trade. Accretion achieved.

Now scale it to the full $28 million. Treasury drops from $130 million to $102 million. Total gross assets fall to $145.5 million. With a reduced share count, per-share coverage could improve by several percent — but the market never prices companies at gross asset value without a discount in a distress scenario. The discount is the market's assessment of the gap between the company's assets and its liabilities, its operating losses, and the risk that the treasury will be drained before creditors are paid. In such a scenario, each successive buyback dollar spent on shares is a dollar that is no longer available for debt repayment, supplier payments, or strategic survival. The market can — and historically does — widen the discount on the stock in response. The company burns $28 million of its last great asset pool, and the stock price doesn't respond at all. That is the bear-market execution risk that cannot be hedged.

This is precisely the kind of loop I dissected during the Terra Luna collapse in 2022. I calculated the exact liquidity depth required to sustain the UST peg and found it mathematically impossible under adverse conditions. The same reasoning applies here: a self-referential scheme — selling assets to buy shares to close a discount that is itself a function of the market's distrust of the asset-selling — is fragile by construction. The mechanism works flawlessly until, at some critical moment, the market decides the buyback is a liquidation in disguise. At that point, the act of selling Bitcoin makes the discount worse. The company ends up selling more treasury to chase a share price that is falling faster than the buyback can catch it.

We chased the glow, not the ledger. The glow is the press release about confident share repurchases. The ledger is the working capital statement that says the firm has more obligations than cash and is now one step away from consuming its most valuable asset to feed the ticker.

THE SYMMETRY WITH STRATEGY AND THE INDUSTRY'S SHIFT

We have seen this note in the industry before. The week before Canaan's filing, Strategy — formerly MicroStrategy — put Bitcoin sales on the table for repurchases. Analysts debated whether a $1.38 billion BTC sale would be digestible if handled off-market. The consensus leaned toward "probably yes, if structured carefully." But the deeper concern was that Strategy's Bitcoin pile had now been explicitly listed as a funding source. Once a treasury is framed as liquidity, its role changes in the market's eyes. It stops being a strategic reserve and starts being a pool of potential supply. That, more than any single sell order, is what pressures Bitcoin's revaluation expectations during a bear cycle.

Canaan's case is Strategy's, but with leverage and without the optionality. Strategy's operating business generates meaningful cash flow; Canaan's generates operating losses. Strategy's treasury is unencumbered at scale; Canaan's has 905 BTC pledged as collateral. Strategy's market cap is substantially larger than its treasury in most valuations; Canaan's market cap is $28.8 million below its gross asset pool. When a healthy company signals willingness to sell BTC, the market discounts the signal as remote. When a distressed company does the same, the market prices the asset sale as probable. The probability, in turn, reduces the value of the remaining treasury, which deepens the discount, which invites more buybacks, which requires more sales. A spiral, not a trade.

CoinShares' mining report, published in late March, identified the industry-wide divergence: stressed miners are selling coins, strong operators are pivoting into AI, and listed mining stocks are becoming less pure Bitcoin proxies. Canaan sits in a dangerous middle lane. It is a seller of hardware, not a mining operator with a huge fleet of its own; it doesn't have enough coins to be a pure treasury story; and its AI pivot prospects are limited by a balance sheet that can't support a major data-center build-out. Its identity is becoming a third thing: a self-funding buyback machine whose source of fuel is the digital asset reserve. That identity is unprecedented, and unprecedented structures deserve extra skepticism.

The company's Q1 results show exactly why. The gross loss of $22.9 million means the company is selling ASICs below their full production cost. The operating loss of $54.3 million means overheads and R&D are adding more fuel to the fire. With Q2 revenue of $35-45 million and a June treasury of $130 million, the question isn't whether Canaan can survive the next two quarters — it can. The question is whether it can survive a scenario where the treasury is drained to $85 million, the share price stays stuck, and the next hardware cycle arrives later than expected. The buyback is a trade-off being made today against a recovery that is not guaranteed tomorrow. It is the kind of decision that looks rational on a loan-to-value basis and acts like a slow-motion liquidation in practice.

WHAT THE ON-CHAIN AUDIT WILL REVEAL

Now the part the financial press won't write. As an on-chain detective, I do not need Canaan's next quarterly filing to know whether it sold. The chain remembers everything. Every block hides a confession.

Canaan's treasury wallets are not anonymous. Over the past two years, blockchain analytics firms have tagged a cluster of addresses connected to Canaan through exchange deposits, mining-pool payouts, and corporate treasury movements. The pattern is traceable: Avalon mining operations send rewards to a consolidation address; the address periodically sweeps funds to major exchanges or OTC desks; and the OTC desks settle in USD, which finds its way back to the corporate bank account. When the SEC filing authorizes treasury sales, the on-chain activity of that wallet cluster becomes the single most important dataset for anyone tracking Canaan's true financial state.

History is written in hex, not headlines. Shareholders reading the Aug. 4 release have permission to look. The moment Canaan moves a significant block of BTC to an exchange address, perceptive on-chain analysts will know. The moment an OTC desk receives a $10 million block of ETH, the market will detect the fingerprint of a mining company converting its reserve. The buyback may be disclosed quarterly, but the sale itself will appear on-chain in real time. Investors who want to understand whether this program is disciplined capital allocation or slow liquidation should be watching the wallets, not the news wires.

I have done this audit before. In 2018, I audited Harvest Finance's early alpha contracts after two weeks of rapport-building with a dev team in Bondi Beach. The social charm opened doors, but the code was the only truth worth trusting. The same principle applies here. Canaan's management is approachable, its brand is known, and its investor-relations materials are polished. But the BTC ledger does not have mood, or sentiment, or optimism. It only has outputs. If the chain shows treasury outflows coinciding with stock buybacks, the strategy is exactly what the filing says it is. If the chain shows outflows that outpace the buyback disclosures, the filing was a mask.

The fixed-income analysts who lent Canaan money against 905 BTC understand this. Their collateral is tracked painstakingly. The equity market, with its quarterly cadence and its tendency to read press releases, is behind. On-chain transparency is the only place where the company's real concentration risk becomes visible: if the price of Bitcoin drops hard, the collateral ratio on those loans tightens, and Canaan must post additional BTC or face liquidation — reducing the available treasury for buybacks and operations simultaneously. That is a cascade risk that no share repurchase program can mitigate. It is, instead, a program that accelerates the exposure.

THE DECEMBER 12 CEILING AND THE TIME CRUNCH

Let me return to the mechanics of the program's timeline. The board set a 12-month window from Dec. 12, 2025. The SEC filing, dated Aug. 4, 2026, effectively expands the funding sources within that window. Roughly 19 weeks remain between the authorization extension and the program's expiry. Canaan has spent about $2 million of the $30 million ceiling; approximately $28 million in capacity remains. To use the full program, the company would need to execute roughly $1.5 million in buybacks per week over the remaining stretch. That is a meaningful pace for a micro-cap with illiquid share trading.

And here is where the time element becomes a mirror of intent. If management genuinely believes the stock is undervalued, the rational move is to accelerate the buyback — front-load the purchases while the discount to gross asset value is steep, and get the greatest per-share accretion. If management is more cautious — worried about liquidity, about the hardware cycle, about loan collateral — it will defer the purchases to month-end windows or tie them to collections milestones. The weekly on-chain treasury flows will reveal which camp management actually belongs to. The Aug. 4 disclosure has created a mechanism; the market must now observe its cadence to infer intention.

The clock also affects the kind of asset sales Canaan would execute. Selling 1,000 BTC over three weeks through public exchanges is a recipe for slippage and front-running. Selling through OTC desks can be discreet but typically requires a price discount of its own. The company has options, but none of them are free. If it sells ETH instead — the relative liquidity of ETH markets, the smaller footprint of its 3,952-token book — it can fund perhaps a quarter of the remaining capacity with modest market impact. The BTC collateral constraints, the loan covenants, and the potential need to maintain a minimum digital asset threshold for lenders all suggest the ETH book is the first to be burned. If the chain shows a small ETH outflow in the first weeks of August, the market will know the buyback has begun. If there are no on-chain movements at all, the disclosure is a signaling tool — a statement of intent designed to justify the stock — rather than a funded operation.

READING THE DISTRESS SIGNATURES IN Q1'S NUMBERS

The deeper I dig into Canaan's Q1, the more the buyback announcement reads like a defensive response to a specific problem: the company's asset coverage ratio has deteriorated to a point where the equity market is valuing it below liquidation value. In Q1, Canaan reported an $88.7 million net loss. Cash dropped by nearly half in a single quarter. Current liabilities exceeded cash by a factor of 2.4. Pledges on the BTC book. A hardware market that CoinShares itself called the weakest cycle for ASIC demand in years.

Every one of these datapoints pushes the market further into valuing Canaan on its liquidation price rather than its going-concern price. A liquidation valuation is brutal: it applies haircuts to inventory, discounts receivables, assumes the mining hardware is sold at auction, and treats the treasury as the only near-par asset. When a company trades at a discount to gross asset value, it is often because the market is calculating a net asset value that the company hasn't shown it can realize. The buyback is an attempt to prove the market wrong by removing shares at a discount, using the very asset the market trusts. If that logic succeeds, the discount closes and the stock re-rates. If it fails, the market treats the buyback as evidence that even management doesn't believe in the operating turnaround — and the discount widens.

This is why the buyback announcement is not neutral to the company's fate. It operationalizes the treasury as a mechanism of price management. And price management, in a distressed equity, is a game with asymmetric downside. When a company with $106.4 million of current liabilities and $43.5 million of cash spends $28 million of its treasury on its own shares, that $28 million is removed from creditor protection. Senior lenders who secured BTC collateral may see their margin cushions shrink. Suppliers may become skittish about extending trade credit. Customers may hesitate to prepay for ASICs if they suspect the company is financing share repurchases rather than manufacturing. The buyback may be the most rational capital allocation on a standalone basis, but it operates inside a web of counterparty relationships that are all watching the same numbers.

Those counterparties are the people who will determine whether Canaan's discount closes or widens. If lenders tighten, the cash burn accelerates. If suppliers demand COD, inventory builds and revenue falls. If customers delay orders, the hardware backlog decays. None of these factors appear in the simple math of the buyback accretion. They are the hidden liabilities that a forensic reader accounts for, and that a press-release reader doesn't.

WHAT THE BULLS GOT RIGHT

Now, because my job is to dissect rather than to sneer, I have to steelman the other side. Cutting against my own skepticism is the requirement of the craft. And the bull case for Canaan's buyback is not without merit. In fact, it has a surprising amount of rigor underneath it.

The first point in its favor is the sheer scale of the discount. A 19.9% gap between market cap and gross asset value is the kind of dislocation that quantitative investors salivate over. Markets are supposed to be efficient; a consistent 20% mispricing is rare and tends to revert. If Canaan closes the buyback at these levels, the remaining shareholders gain real per-share value. The treasury's composition — roughly three-quarters Bitcoin, a quarter Ethereum — acts as a built-in hedge against fiat depreciation. Selling a small portion of that treasury to lock in a cheap equity price is, in isolation, a sophisticated capital-market maneuver. An analyst looking at this trade in a vacuum would say: buy the stock, because the company is buying itself at a wickedly good price.

The second point is the optionality of the SEC filing's vagueness. The fact that Canaan disclosed no new sales and no new repurchases as of the Aug. 4 release means the authorization is, for now, purely a tool. Management can choose never to use it. It can choose a single small sale to fund a token buyback, enough to register confidence without depleting the treasury. The mere existence of the authority may stabilize the share price — because the market now knows there is a floor of potential demand for the stock. That's a cheap form of price support: no BTC sold, no shares bought, but the disclosed capacity itself functions as a market signal. And in a bear market, signals matter.

The third and most contrarian point: the industry data supports the trade. CoinShares' report noted that the biggest shift in mining is the bifurcation of behavior — the weak sell BTC to survive, the strong sell equity to fund AI pivots. Canaan, with its hardware business, is actually in a position to benefit from the "picks and shovels" recovery if the cycle turns. ASIC demand is cyclical; it collapsed now, but it will recover when Bitcoin production economics improve. If Canaan can hold its hardware market share and maintain its treasury through the downturn, the combined asset base is a leaner, stronger business than the current price reflects. The buyback is not a retreat; it is an opportunistic acquisition of its own equity at fire-sale prices.

I have to concede that each of those arguments has merit. Markets do overcorrect during periods of distress. Cheap buybacks do create long-term value when the buying price is below intrinsic value. A company with an $88.7 million quarterly loss can still be a good investment at the right price — because the price embeds the assumption of continued losses for years. At $1.50 per ADS with $1.83 of gross asset coverage per share, the market is pricing in a permanent operational decline. If management believes the hardware cycle is due for an upturn, the fair value of the going concern is materially higher than the liquidation value. Capturing that gap by purchasing stock is, in the truest sense, an investment thesis.

So the bull case is not delusional. It is simply conditional. It depends on three assumptions: that the discount is driven by market sentiment rather than by known liabilities; that the treasury can be sold without impacting the asset price; and that the operating business does not require the full treasury to survive the next cycle. Each of those assumptions is testable, and at least one of them is likely to fail.

TAKEAWAY

The filing was not a strategy. It was a symptom. Canaan's management, facing a business that burns cash in a dying hardware cycle, has made a choice: the Bitcoin and Ethereum it mined, accumulated, and pledged will now be permitted to retire at the altar of its own share price. Minted in hope, burned in regret. I have read this playbook in Treasury bonds, in DeFi protocols, and in algorithmic stablecoins. It always begins with a rational-sounding mechanism and ends with a balance sheet that no longer has a reserve to fail into.

The honest question for every Canaan investor is not whether the buyback is accretive on paper. It is whether the treasury is a strategic asset or a cushion. If it is a strategic asset, selling it to buy equity is a transfer of conviction from one asset to another, and the market will eventually re-rate. If it is a cushion, spending it on buybacks is the equivalent of a parachute being opened before the airplane leaves the runway. The chain will render the verdict long before the next earnings call. I know this the way I know the exact math of Terra's failure: because I looked at the liquidity requirements and found them impossible. Look at Canaan's working-capital requirements and tell me the buyback is free. It is not. Liquidity flows, but integrity stagnates. Every block hides a confession. This one is on the balance sheet, signed, sealed and about to be sent to the chain.