Between the blocks, silence screams the truth.
When a company built on a ‘never sell’ pledge begins liquidating its core asset to fund a dividend, the market’s first reaction is denial. The second is a frantic search for confirming signals. The Q2 13F filings for Strategy (formerly MSTR) delivered exactly that signal: 12 of the top 15 institutional holders increased their positions. Net institutional inflow: $700 million.
But that number is a mirage. The structure beneath it tells a very different story—one of a capital machine quietly consuming its own fuel.
Context: The Machine That Was Promised to Run Forever
Strategy’s model is pure financial engineering. The company raises equity or preferred stock (STRC), buys Bitcoin, and uses the Bitcoin holdings to back its own stock price. The flywheel: higher BTC price → higher NAV → higher stock price → more capital raised → more BTC bought. The implicit promise was ‘never sell Bitcoin.’ That promise was the foundation of the premium over Bitcoin ETFs. In 2022, during the depths of the bear market, I analyzed MSTR’s reserve structure for a quantitative fund. The model held because the thesis was binary: either Bitcoin collapses or it appreciates. There was no third state—no periodic cash drain.
That changed in May 2026. Strategy began selling Bitcoin to fund STRC preferred stock dividends. The conversion from a ‘buy-and-hold’ treasury to a ‘buy-and-consume’ vehicle is not a tweak. It is a structural mutation.
Core: The Data That Unglues the Narrative
Let me walk through the on-chain and regulatory evidence chain.
First, the sell pressure is real and recurring. Since May, Strategy has executed multiple Bitcoin sales. The stated purpose: covering STRC dividend obligations. This is not a one-time rebalancing. The STRC preferred stock carries a fixed dividend schedule. If the company has no other operating cash flow (it doesn’t), those dividends must be paid from Bitcoin sales. Every quarter, the market faces a new forced sell order—regardless of price.
Second, the institutional buying is structurally deceptive. The Q2 13F data shows 12 of 15 top holders increased. But look at the composition:
- Vanguard (two entities) added $147 million combined.
- BlackRock Institutional Trust added $84 million.
- Capital International (active) increased, but not by a disclosed amount.
- Goldman Sachs nearly quadrupled its position to $555 million.
Meanwhile, Capital Research Global Investors—a large active manager—dumped $462 million. UBS cut $142 million. Geode trimmed $5 million.
The net is positive, but the direction is split. Passive funds (Vanguard, BlackRock) are index-rebalancing. They cannot choose to sell if the stock remains in the index. Their buying is a mechanical artifact of allocation rules, not a vote of confidence in Strategy’s broken promise. The active manager exodus, however, is a vote. Capital Research Global Investors alone accounts for 76% of all selling volume among the top 15. That is a signal.
Third, the Goldman Sachs position deserves scrutiny. I have worked with prop desk data. A $555 million position that ‘nearly quadrupled’ is often a hedge or a market-making book, not a long-term strategic allocation. Goldman’s desk might be using MSTR to arbitrage the premium/discount versus Bitcoin futures. That is not endorsing the company’s management—it is exploiting the structure.
Fourth, the velocity of the consumption flywheel is accelerating. In Q1, net institutional additions were $4.6 billion. In Q2, $700 million. That is an 85% collapse in marginal demand. Meanwhile, the dividend obligation is fixed. The math is simple: if new capital inflows slow while outflows (dividends) continue, the delta must be filled by selling more Bitcoin. Floors are illusions until you map the liquidity.
Contrarian: The Passive Fund Mirage
The conventional read is ‘12 out of 15 institutions increased, so confidence is intact.’ That is a correlation trap. Let me break the causality.
Passive fund buying is not a signal of conviction; it is a signal of index inclusion. Vanguard and BlackRock collectively added $231 million. If the market cap of MSTR fluctuates, passive funds mechanically adjust to maintain weight. Their Q2 buying could have been triggered by a drop in MSTR’s relative index weight after Q1’s massive inflows. In other words, they bought because the stock got cheaper in the index, not because they believe in the Bitcoin treasury strategy.
Active managers, on the other hand, have discretion. Capital Research Global Investors cut $462 million. That is a clear statement: the risk-reward of holding MSTR, given the new ‘sell to pay dividends’ reality, no longer justifies the exposure.
Another blind spot: the STRC preferred stock itself. The dividend is fixed. If Bitcoin price stays flat or declines, the company must sell more BTC to cover the same dollar obligation. This is a negative convexity. In a downturn, the sell pressure rises. This is not a risk that ETF holders face. ETFs do not have to sell Bitcoin to pay dividends. They simply hold. The structural advantage of ETFs over MSTR is widening.
Structure creates freedom; chaos demands order. The current structure of Strategy is chaotic: a fixed liability in a volatile asset. The market has not yet priced this because the passive fund flows are masking the deterioration. But when the next 13F filing comes, the noise will fade, and the signal will be unmistakable.
Takeaway: The Next Quarter Is the Test
If Q3 13F shows a continued decline in net institutional additions, or if the active-to-passive ratio tips further toward selling, the premium will collapse. At that point, the flywheel reverses: Bitcoin sales to cover dividends reduce NAV, which reduces the stock price, which reduces the ability to raise new capital, which forces more Bitcoin sales. The floor is unknown until we map the liquidity demands of STRC holders.
Watch the next 13F. Watch the Bitcoin sales frequency. The silence between the blocks is already screaming.