The Bitcoin Treasury Model Is Splitting. MSTR’s Leverage Is the First Casualty.

The vote passed with over 90% approval. Satsuma Technology, a pure-play bitcoin treasury, is liquidating its 668 BTC. No drama. No hack. Just a shareholder referendum on a model that stopped working. The date: July 21, 2026. The signal: pure holding without cash flow is dead. I’ve spent the last decade auditing balance sheets disguised as protocols. The Satsuma liquidation isn’t an isolated failure. It’s the market test that every bitcoin treasury — including Strategy, Inc. (MSTR) — just failed. The only question is how long MSTR’s credit model can outrun the same verdict. VanEck’s Matthew Sigel says many pure bitcoin treasuries are exiting accumulation strategies. That’s a polite way of saying the “buy and hold BTC forever” thesis cannot survive public market scrutiny. Satsuma’s 668 BTC — roughly $43.5 million at current prices — is being returned to shareholders. The model’s “exit” is now a line item in a liquidation statement. Here’s the structural problem. A bitcoin treasury company has one asset: BTC. It has zero operating cash flow. Every dollar of shareholder value depends on either bitcoin appreciation or someone else paying a premium for the stock. When the premium disappears — and it always does — the company becomes a discount vehicle with no floor. Satsuma learned this empirically. What’s painful is that MSTR is running the same playbook with a leverage overlay. I’ve written before about the dangers of reflexivity in DeFi. MSTR’s credit model is the corporate equivalent of a leveraged token: issue convertible debt, buy BTC, watch the stock rise with BTC, use the higher stock price to issue more debt. The flywheel works in a bull market. The 2021-2025 cycle gave Saylor cover. But the machinery underneath is a chain of payment obligations with no underlying production. It’s not malicious. It’s just fragile. The liquidation of Satsuma coincides with the emergence of two successor models, announced about a week apart. First, Orange Juice — a permanent capital vehicle backed by Lyn Alden and Jeff Booth. Second, Twenty One Capital — a Tether-supported entity. Both are designed to hold BTC forever, but with a twist: they claim to generate cash flow. That’s the new narrative. But I’m a forensic analyst, not a jpeg collector. I need to see the capital structure. Let’s dissect the models. The purist model: hold BTC, do nothing. That’s Satsuma. No income, no product, no hedge. The treasury is literally a cold wallet with a ticker. As a math exercise, the net present value of future cash flows is zero. If bitcoin stagnates, this entity has no way to pay salaries, audit fees, or even the coffee bills. The only “profit” is unrealized BTC appreciation, which no creditor accepts. Liquidate. Next. The credit model: MSTR. Buy BTC with debt and equity. The cost of capital is the risk-free rate plus a credit spread and the dilution cost of issuing shares. In exchange, shareholders get a leveraged bitcoin exposure. In a bull market, this leverage amplifies BYX gains. In a flat or declining market, the premium decays. The model depends on a continuous stream of new buyers at high prices. When the premium closes, the arb reverses: short MSTR, long BTC. I’ve seen this exact pattern in every convertible bond landmine I’ve audited. But there’s a subtle accounting detail most retail holders ignore. MSTR’s bitcoin holdings are carried at fair value under ASC 350, but the associated liabilities — the converts, the preferred shares — are marked to yield. The duration mismatch between a perpetual asset (BTC) and a finite-maturity liability (debt) creates a solvency cliff. If BTC drops 50%, the equity goes negative even though the bitcoin still exists. That’s not a black swan. That’s a defined risk. Entropy wins. Always check the fees — and the maturity dates. The permanent capital model: Orange Juice, Twenty One Capital. The promise is a Berkshire-style structure: own BTC, generate income via lending, arbitrage, or other treasury operations, then reinvest that income into more BTC. Sounds elegant. In practice, it’s an unregulated investment fund with a bitcoin kicker. Governance becomes paramount. Who decides the lending rate? Who audits the collateral? What happens if the lending counterparty collapses? I spent four months reverse-engineering FTX’s withdrawal engine after 2022. The same opacity that killed them can infect a permanent capital vehicle. The article hints that Twenty One Capital, backed by Tether, may have a structural advantage: access to the USDt issuance float. Tether generates billions in profits from reserve yield. If Twenty One Capital uses that cash flow to buy BTC, its cost of capital is materially lower than MSTR’s convertible debt. That is a competitive weapon MSTR cannot match. I’ve seen this asymmetry before — in Layer 2s where one player controls the sequencer and everyone else rents gas. The centralization of capital is more dangerous than any code bug. Now, the contrarian take. Most analysts will frame this as “MSTR must evolve or die.” I disagree. The shift is not about sophistication. It’s about exit liquidity. Satsuma’s liquidation is a dignified exit. MSTR’s leverage is an accidental one. The new permanent capital vehicles are not solutions; they’re experiments with unproven governance and unknown liquidation terms. What’s actually happening is a market-wide repricing of bitcoin treasuries from “store of value collectives” to “cash-flow-producing financial companies.” That’s a category change. The public will demand EBITDA. If the treasury can’t produce it, the stock trades at NAV discount. And that is where the real risk sits. Consider the hidden signal in Satsuma’s vote. 90% approval means the shareholders weren’t waiting for a rescue. They understood the model had no future. That rational consensus is what MSTR lacks today. MSTR’s shareholders are still paying a premium because they expect the leverage to pay off. When that expectation breaks, the premium will collapse — not gradually, but as a gap down, because the shareholders who remain are the most leveraged and the most sensitive to any dip. I ran a simple simulation for my own portfolio. Based on my work simulating EIP-1559 fee markets, I modeled MSTR’s equity value as a call option on BTC with a strike price at its debt obligations. With current debt levels and a 30% BTC drawdown, the call goes deep out of the money. The stock doesn’t just fall; it underperforms bitcoin by the leverage ratio. The credit spread widens, the convertible holders get a floor, and the equity becomes a lottery ticket with negative skew. That’s not fear. That’s math. The permanent capital model isn’t safe either. A “forever” company still has to fund operations. If it issues zero-dividend shares, the only return is NAV growth. But if it uses lending to generate yield, it inherits counter-party risk and liquidity risk. The market’s appetite for these structures will depend on the same metrics I’ve used for years: cost of capital, recovery rate, and liquidation mechanics. The absence of these data in the announcements is a red flag. Let me be precise. The Bitcoin treasury model is shifting because the era of free money ended. The pure hold model was a product of a rising market. The credit model is a creature of a leverage-friendly regulatory window. The permanent capital model is an attempt to build a coasting sailboat in a hurricane. Each model is a different version of the same trap: believing a balance sheet can replace a business. From my ZK-Rollup audit days, I learned that every recursive proof has an edge case. The counterparty to those edge cases is human trust. The same applies here. The only sustainable treasury model is one that generates value beyond the mere custody of bitcoin. That means real operations, real accounting, and a real obligation to minority shareholders. Anything less is a house of cards. Seven days after Satsuma’s vote, Orange Juice and Twenty One Capital announced. They will likely attract capital because the alternatives look worse. But the clock is ticking. If these entities cannot prove positive cash flow within two quarters, the market will move on rather than wait. 2017 vibes. Proceed with skepticism. I’ve seen this before — protocols that promise “real yield,” then dissolve when the token price drops. The difference here is that the underlying asset has intrinsic value. Bitcoin is not going to zero. But the vehicles that hold it can go to zero. The leverage is the liability. The liquidity premium is the mirage. For MSTR specifically, the floor is the bitcoin per share value. If it trades below that, arbitrage closes the gap — but only if the market can short the stock freely. With high borrow costs, the discount can persist for months. That persistence is what causes margin calls and forced liquidations. I’d watch MSTR’s implied volatility and borrow fee as leading indicators, not the price. My takeaway is not “sell MSTR.” It’s “understand the fee structure.” Every capital model has an embedded fee: the interest on convertible debt, the spread on preferred shares, the carry cost of a permanent capital vehicle. That fee is the price of leverage. When the fee exceeds the realized volatility of bitcoin, the model is underwater. Right now, for MSTR, that crossover is closer than the market thinks. The next 12 months will separate the financial engineers from the carnival barkers. Satsuma’s liquidation is a clean exit. Orange Juice and Twenty One Capital are clean starts. MSTR is the stress test. And as I’ve learned from every audit, the failure mode is always in the assumptions about liquidity. Entropy wins. Always check the fees. And before you buy any treasury stock, ask one question: what is this company’s actual cost of carrying its bitcoin? If the answer is not a number, you’re the exit liquidity.