The Discount That Isn't There: Canaan's Buyback Lever and the Liquidity Trap

Liquidity is a mood, not a metric. On August 4, a terse SEC filing from Canaan captured the current mood of the crypto mining industry with uncomfortable precision: management may now use proceeds from the company's roughly $130 million digital asset treasury to fund share buybacks. The market barely blinked. No new treasury sale was disclosed. No new repurchase was confirmed. The filing simply opened a door — and attached no execution timeline, no asset amount, and no indication of whether the lever would ever actually be pulled. But that lever moves more than the mechanics of a buyback. It moves the story of how a hardware maker in a downcycle treats its assets, and whether the market's refusal to fully credit the treasury is a failure of perception or a rational reading of a balance sheet under pressure. The gap between the headline numbers and the operational reality is where the truth sits. That gap is also where the easy narratives get built. Let me lay out the timeline first, because sequence matters here. The buyback program began on December 12, 2025, with a 12-month ceiling of $30 million for American Depositary Shares or Class A ordinary shares. By May 19, Canaan had spent approximately $2 million to repurchase about 2.8 million ADSs, according to its first-quarter results. Simple subtraction puts nominal remaining capacity at approximately $28 million as of that date — though the August 4 release declined to disclose where the newer repurchase total stands, so the true unused authorization remains, for now, a known unknown. Meanwhile, the treasury grew. The company's June operating update listed 1,915 BTC and 3,952 ETH on the balance sheet as of June 30, including receivables and excluding customer deposits. At early August prices, that digital asset position sits at roughly $130 million. Combine it with the separately reported March 31 cash balance of $43.5 million, and the gross sum of cash plus crypto reaches approximately $173.5 million. Against the company's intraday market valuation of $144.7 million on August 4, sourced to S&P Global Market Intelligence via StockAnalysis, the difference is approximately $28.8 million — or 19.9% of the market cap. This is the number that will circulate. It seems to suggest that Canaan's equity trades roughly a fifth below the value of its most liquid assets, and that the new buyback channel is a rational response to a market that refuses to see value. I have spent enough time inside crypto balance sheets to know that treasuries rarely behave the way headline valuations suggest. In the summer of 2020, I spent forty hours manually tracing $2.5 million in USDC flows from Compound to Uniswap V2 for my thesis on monetary policy transmission. What that exercise taught me is that balance-sheet surpluses in crypto are never as simple as the marks suggest. Hiding inside them are lockups, pledges, counterparty exposures, and the slow realization that not every token is a liquid asset in a storm. So let me do the subtraction that the headline framing omits. The $173.5 million figure combines a digital asset estimate at August 3 prices with a cash balance from March 31. The market cap is intraday on August 4. This is not a same-date net asset value — and more importantly, it is not net of liabilities. The inputs come from different dates, and any directional view of a discount needs to confront the rest of the balance sheet before it becomes an actionable thesis. Canaan's March 31 balance sheet shows $43.5 million in cash against $106.4 million in current liabilities, down from an $80.8 million cash position at year-end. The company had also pledged 905 BTC as collateral for secured term loans and placed another 100 BTC into a fixed-term product. These figures predate the June holdings, but they draw the critical distinction: the headline treasury and the assets actually available for near-term deployment are not the same pool. Something between those two numbers houses the company's real capacity to survive. Here is the exercise I run with my own models. Take the gross asset sum of $173.5 million. Subtract the known current liabilities of $106.4 million. The result is approximately $67 million in partial net asset value — and that is before considering long-term debt, operational losses, tax positions, or the unknown quantity of restricted crypto. Compare that $67 million to the $144.7 million market cap, and the narrative inverts. The equity is trading at more than double its simple net asset value, not at a discount. The "20% gap to market cap" that the August announcement seems to gesture toward is not a mispricing. It is a mirage generated by comparing gross assets against equity value — an apples-to-oranges construction that looks cheap on first glance and expensive on second. A working capital view makes the tradeoff even more acute. Canaan reported a $22.9 million first-quarter 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 came in line with the lower band of prior forecasts, which reads more like containment than recovery. The weakened hardware cycle makes the expanding treasury increasingly central to how Canaan is valued — but it also makes every dollar minted from that treasury more consequential. Now consider what the buyback authorization actually does. Each dollar directed toward repurchases is a dollar not spent on servicing the $106.4 million of current liabilities. The crypto treasury, effectively the only large pool of value the company controls that generates no operational drag, becomes both the survival fund and the shareholder returns fund. Two independent claims on one pool is a fragile structure. Structure is the skeleton; liquidity is the blood. But blood can only flow to one organ at a time. When I collaborated with senior portfolio managers in Warsaw in March 2024 to model the inflow of institutional capital after the first spot Bitcoin ETF approvals, one lesson kept surfacing: the hardest variable is never supply and demand, but velocity — how quickly assets marked on ledgers become assets moving through markets. For passive fund flows, the assumption is mechanical. For operating companies with pledged collateral and managerial discretion, it never is. The market knows this. The market prices this. And that is why the discount narrative fails under scrutiny. The contrarian position here is uncomfortable: the market may be right. Illusions fade when the tide of liquidity recedes. Anyone who sees a 20% discount in Canaan's structure is buying a stat that gives no weight to the liabilities, the pledges, or the operating burn. The discount is not there once current liabilities enter the frame. The company is valued not against the gross value of its resources, but against its ability to survive the hardware downcycle while meeting obligations. Market caution in the face of a $88.7 million quarterly net loss is not a failure of perception. It is a rational response to a deteriorating operational baseline. The mechanism itself deserves scrutiny too. Using crypto proceeds for buybacks transfers value from one constrained pool to shareholders while the core business continues consuming cash. I examined similar dynamics during my January 2025 audit of five staking providers ahead of the EU's MiCA implementation, watching $500 million in staked assets get reclassified as securities and, in some cases, become restricted collateral. The lesson stuck: when a filing touts the flexibility of balance-sheet assets, it is often a signal that the operating business cannot generate the liquidity it needs on its own. Canaan is not alone in this pattern. Strategy has put Bitcoin sales on the table for repurchases. Stressed miners are selling coins to stay liquid while stronger operators pivot toward AI infrastructure. The industry is quietly shifting from treasuries as conviction assets to treasuries as funding sources. This is the macro mirror of the micro decision: when the most visible balance sheets treat bitcoin as a liquidity buffer rather than a strategic reserve, sentiment has shifted at the margin — even if the spot price continues to say otherwise. The macro is the mirror of the micro. The market will learn the true intent only through execution. The filing authorizes a lever; it does not confirm a pull. Watch the treasury sales, the scale of each subsequent buyback, and the monthly operational updates for clues. What matters is not that Canaan has the option to buy back stock with crypto. What matters is whether, after the liabilities, the pledges, the restricted coins, and the operating losses, there is any liquidity left to spend at all. Patterns repeat, but the context never does. In 2022, the crash stripped away every company that tried to fund growth with borrowed confidence. Today, we watch to see how many companies fund stability with borrowed liquidity. Canaan's future is written not in the $28 million authorization that nominally remains, but in the present liquidity of its core business. That is the number to track. The cycle will find its rhythm again, and when it does, the question will not be whether Canaan repurchased stock at a discount. It will be whether that stock still exists.