Tracing the code back to its genesis block – except this time, the code is not smart contracts but SEC-registered preferred stock. Late last week, Strategy (formerly MicroStrategy) disclosed that it had raised approximately $15 billion in cumulative capital through two novel preferred stock instruments, STRK and STRC, with the design process accelerated by an AI model. The news barely moved the bitcoin price, but the structure itself is a cryptographic artifact of financial engineering that deserves a forensic dissection.
Context: The Saylor Doctrine Michael Saylor’s transformation of a legacy software company into the world’s largest corporate bitcoin holder is well documented. What is less understood is the financial machinery that funds it. After exhausting the traditional playbook – common stock dilutions (ATMs) and zero-coupon convertible bonds – Saylor faced a bottleneck. The market’s appetite for unsecured convertible debt was finite, and the cost of equity was too high.
Decoding the signal hidden in the noise – the noise here is the hype around "AI-designed securities." The reality is more mundane and more fascinating. Strategy needed to tap into a different investor base: fixed-income funds, pension plans, and yield-seeking retail. The solution was a two-tier preferred stock structure: STRK, a fixed-rate convertible (10% dividend, convertible into common shares), and STRC, a floating-rate non-convertible (dividend adjustable, price anchored near $100 par). The latter is essentially a credit instrument – a short-term loan to the company secured by the promise of future bitcoin appreciation.
Core: The Mechanics of Leverage Let’s follow the smart contract, ignore the whitepaper. The financial engineering works as follows:
- STRK – Fixed dividend of 10% annually. Investors receive the coupon and the option to convert into MSTR common stock, which is a leveraged play on bitcoin. The cost of capital to Strategy: 10% pre-tax.
- STRC – Floating dividend, initially set at 6.6% but adjustable. The price is intentionally kept near $100 to signal stability. This is not a debt instrument, but it behaves like one: the company pays dividends from its operating cash flow or from new issuance. The total funding from STRC alone is approximately $10.5 billion, with another $4.5 billion from other preferred securities, including STRK.
Where liquidity flows, truth eventually pools. The truth is that Strategy is selling credit – not equity. Saylor himself admitted in the podcast: "We basically sold $15 billion of credit." The AI model, which was used to explore structural designs and regulatory boundaries, did not create the credit; it merely optimized the packaging. The real innovation is the dividend adjustment mechanism: if market rates rise or bitcoin sentiment sours, Strategy can increase the STRC dividend to retain investors. In a bull market, this is a powerful tool to keep funding cheap. In a bear market, it becomes a drain on cash flow.
Composability is a double-edged sword. The composability here is between the company’s bitcoin stack (now over 840,000 BTC) and the financial markets. Each new issuance of STRC adds to the asset base, but also to the fixed dividend obligations. The net effect is a leveraged long position on bitcoin, funded by a combination of fixed and floating rate liabilities. The cost of leverage is approximately 7-10% annually, depending on the mix. If bitcoin’s average annual return exceeds that, common shareholders win. If not, the company faces a cash-flow crisis.
Contrarian: The AI Narrative is a Distraction The contrarian angle is not that the instruments are flawed, but that the AI narrative is a deliberate misdirection. Saylor understands that the market rewards innovation stories. By attributing the design to AI, he positions Strategy as a tech-forward company rather than a leveraged bitcoin fund. This is a narrative play, not a technological breakthrough.
Bubbles burst, but architecture remains. The architecture of STRK/STRC is sound within the boundaries of U.S. securities law, but it relies on a single assumption: that bitcoin will continue to appreciate over the long term. If bitcoin enters a multi-year bear market, the dividend payments will accumulate, and the company will be forced to either sell bitcoin (defeating the purpose) or issue more debt at higher rates – a classic debt spiral. The AI model did not account for this tail risk; it was trained on historical data that includes only one major bitcoin bear cycle.
Furthermore, the investor base for these instruments is not as sophisticated as institutional buyers of traditional preferred stock. Many retail investors are attracted by the fixed yield and the “bitcoin backing” without fully understanding that the dividend is paid from the company’s cash flow, not from bitcoin itself. In a downturn, the $100 price anchor could break, leading to panic selling and a collapse of the funding channel.
Takeaway: The Next Narrative The takeaway is not to dismiss Strategy’s approach, but to recognize it as a product of the current cycle. The next narrative will be whether other companies replicate this model, and whether the SEC will scrutinize the use of AI in financial product design. For now, Saylor has built a financial machine that works perfectly in a bull market and amplifies risk in a bear market. The question is not whether the code is solid, but whether the market will remain hospitable. As the saying goes: follow the smart contract, ignore the whitepaper. Here, the smart contract is the dividend adjustment clause, and the whitepaper is the AI story. One is real, the other is noise.
