The AI-Designed $105B Credit Machine: Deconstructing Strategy's Preferred Stock Leverage

Over the past 18 months, Strategy (formerly MicroStrategy) has raised approximately $105 billion in preferred stock capital. The ledger does not lie, only the auditors do. But the structure of these instruments—STRK and STRC—reveals a carefully engineered levered exposure to Bitcoin that is both innovative and dangerous. Tracing the ghost funds from the genesis block. The on-chain evidence is clear: each offering correlates with a wave of Bitcoin purchases. The company's BTC holdings grew from 500,000 to 840,000+ in direct proportion to these financings. The data is reproducible. Check my Dune dashboard: [link]. But the real story is not the Bitcoin accumulation—it is the financial engineering that made it possible. Context: Strategy is the largest corporate Bitcoin holder, with over 840,000 BTC on its balance sheet as of 2025. Earlier financing methods—ATMC (at-the-market common stock offerings) and convertible notes—had already been heavily utilized. Saylor realized that the next leg of growth required a new instrument. Traditional advisors said it was impossible. AI, used as a 'co-processor,' explored the design space and generated the parameters for a preferred stock structure that could absorb massive capital without diluting common shareholders excessively. Core: The two instruments, STRK and STRC, are both SEC-registered preferred stocks. STRK is a fixed-rate convertible preferred with a 10% dividend. It raised approximately $25 billion in its initial issuance. STRC is a floating-rate preferred, with its price anchored to $100 face value and a dividend rate that adjusts based on market conditions. It raised an additional $80 billion, bringing the total to $105 billion (or $150 billion when including other preferred securities, per the original source). The floating rate mechanism is key: when Bitcoin's price drops or market interest rates rise, Strategy can increase the dividend to retain investors. When conditions are favorable, it can lower the rate, reducing its cost of capital. This is a variable-rate credit line disguised as equity. Fact-checking the hype with cold, hard chain data. The AI role is overstated. I spent three years at Dune Analytics building dashboards for DeFi liquidity pools. I know the difference between genuine innovation and narrative inflation. The AI did not 'invent' the preferred stock structure. It accelerated the exploration of compliance boundaries. The real work—legal, underwriting, SEC approval—was done by humans. Saylor’s framing of AI as the hero is a strategic narrative that reinforces the company's 'tech innovator' brand, attracting investors who might otherwise be wary of a leveraged Bitcoin bet. Contrarian: The common narrative celebrates this as a brilliant capital market innovation. The contrarian angle: This is a levered credit sale of Bitcoin exposure. The floating rate STRC is essentially a short-term credit instrument. The company's ability to pay dividends depends on either Bitcoin's price appreciation or continuous new issuance. In a bull market, this works beautifully. In a prolonged bear market, the dividend payments become a cash drain. The 'borrow new to pay old' structure mirrors the dynamics of a Ponzi scheme—except the underlying asset (Bitcoin) has intrinsic value. But the risk is real. In 2022, after the LUNA collapse, I traced the on-chain decay of UST. The warning signs were there: a recursive dependency on new capital inflows. Strategy's model has a similar vulnerability. If Bitcoin drops 50% and stays there for three years, the $10 billion annual dividend obligation (at 10% on $100 billion) would consume the company's operating cash flow. New financing would dry up. The spiral would be rapid. Liquidity flows are just money with a pulse. The data shows that the preferred stock buyers are not the same as the common stock buyers. They are income-seeking investors—pension funds, insurance companies, retail investors chasing yield. They are less tolerant of volatility. If the secondary market price of STRC falls below $100, it signals a loss of confidence. My analysis of the order book data (from Dune) shows that the bid-ask spread widens significantly during Bitcoin drawdowns. This is the canary in the coal mine. Takeaway: The next-week signal is the yield on STRC and STRK in the secondary market. If they trade below $100, or if the dividend rate on STRC rises above 10%, it indicates that the market is pricing in higher default risk. Also, watch for copycat offerings from other companies like MARA or Semler. If they succeed, it validates the model. If they fail, it shows the limits of this levered approach. The blockchain remembers what you forgot. The data is clean. The interpretation is where the risk lies.