Strategy bought 4,603 bitcoin for $369.7 million between August 24 and 30, its first purchase since June and a sharp reversal after ten straight weeks of selling. The buy pushes the company's treasury to 845,050 BTC, and it lands while Strategy's own stock still trades at a discount to the value of that pile, a state the market rarely put Michael Saylor's company in before this year.

The company paid an average of $80,318 per coin, funded by selling 4.53 million shares of MSTR through its at-the-market equity program for $602.8 million in net proceeds. That's the mechanism worth sitting with for a second: Strategy didn't dip into cash reserves or take on new debt for this one, it printed stock and used the proceeds to buy bitcoin, the same playbook that built the company's 845,050-coin position over the past five years.

RelatedStrategy Sold Bitcoin Below Cost to Prop Up Its Preferred Stock

Why did Strategy stop buying in the first place?

Since June 22, Strategy had been a net seller, offloading roughly 7,000 BTC for about $432.5 million. That's unusual for a company whose entire public identity is "buy bitcoin, never sell," and the reason wasn't a change of conviction on bitcoin itself. It was balance-sheet plumbing: Strategy needed to fund dividend payments on its stack of preferred shares (STRK, STRF, STRC and friends) and cover a share repurchase program, and selling a sliver of the treasury was cheaper than raising new capital at a moment when MSTR stock had fallen well below the value of its bitcoin holdings.

That gap between stock price and treasury value is measured by something called mNAV, modified net asset value, and it's the number that explains almost everything about Strategy's behavior this year. For most of its bitcoin-buying history, MSTR traded at a premium to mNAV: investors paid more for a share of Strategy than the bitcoin behind it was worth, essentially betting on Saylor's ability to keep growing bitcoin-per-share through clever capital raises. That premium collapsed in 2026. By late August, Strategy's diluted mNAV sat near 0.74, meaning the stock traded at roughly a 26% discount to its own treasury.

Strategy's net bitcoin flow, before and after August 24 Bar chart showing Strategy sold about 7,000 BTC for $432.5 million during a 10-week pause from June 22 to August 23, then bought 4,603 BTC for $369.7 million between August 24 and 30, reversing course. NET BITCOIN FLOW · STRATEGY (MSTR) 845,050 BTC total treasury SOLD -7,000 BTC $432.5M · Jun 22-Aug 23 BOUGHT +4,603 BTC $369.7M · Aug 24-30 genztech.blog
Fig 1 Ten weeks of net selling to fund preferred dividends, reversed in a single week once fresh equity capital came in.

What actually changed this week?

Two things moved at once. Bitcoin itself firmed up, briefly touching $81,000 on August 25 before settling into the high $70,000s, enough to stop the bleeding on Strategy's mNAV without fully closing the gap. And Strategy found a cheaper way to raise money: instead of selling bitcoin, it went back to its at-the-market share program, selling new MSTR stock directly into the market and using that cash to buy coins. Selling stock at a discount to your own treasury sounds like a bad trade, and in a strict sense it is, existing shareholders get diluted at a below-par price. But it beats selling the underlying bitcoin outright, which permanently shrinks the one asset the whole company is built around.

  • 4,603 BTC bought for $369.7 million, average price $80,318 per coin
  • Funded by selling 4.53 million MSTR shares for $602.8 million net
  • Ends a 10-week stretch as a net seller that moved about 7,000 BTC
  • Treasury now stands at 845,050 BTC, still the largest corporate bitcoin holding on earth

Why does it matter that Strategy is buying below its own NAV?

Because it inverts the logic that made Strategy's model work in the first place. When MSTR traded at a premium to its bitcoin, issuing new shares to buy more bitcoin was accretive: each new dollar of stock bought more than a dollar of bitcoin per existing share. At a discount, that math runs backward. Raising $602.8 million by selling equity worth less than the bitcoin it represents, then using that cash to buy bitcoin, still grows the absolute BTC count, but it does so at the expense of existing shareholders' claim on that stack. Saylor is choosing treasury growth over per-share optimization right now, a bet that the discount is temporary and that scale still matters more than precision timing.

RelatedStrategy's mNAV Falls Below 1, Forcing a Bitcoin Pivot

What it means for the stock

MSTR is the exposed name here, and the signal for investors is mixed rather than clearly bullish. The resumption shows management still has appetite and access to capital even in a down mNAV environment, which matters for anyone worried Strategy might be forced into a fire sale. But the method, dilutive share issuance at a discount, is not free: it confirms the company no longer commands the premium that let it grow bitcoin-per-share for free in 2024 and 2025. Watch mNAV itself more than the headline BTC count going forward. A sustained move back above 1.0 would validate this week's bet; a deeper slide toward the low 0.7s would put pressure on the preferred-share dividend math that forced the June selling in the first place.

  1. Jun 22Strategy begins selling bitcoin to fund preferred dividends and buybacks
  2. Aug 10Confirmed selling below cost basis on some tranches, a first for the company
  3. Aug 19MSTR jumps 14% in a day as mNAV briefly recovers to 1.05x
  4. Aug 25Bitcoin touches $81,000 before easing back toward the high $70,000s
  5. Aug 24-30Strategy buys 4,603 BTC for $369.7M first net purchase since June
What to watch
  • mNAV trajectory. If it climbs back over 1.0, expect Strategy to lean harder into equity-funded buying again.
  • Preferred dividend coverage. Another cash crunch there would be the clearest signal this pause-and-resume cycle repeats.
  • Whether other treasury companies follow. Strategy's peers copying its model, from Metaplanet to smaller US-listed bitcoin holding companies, tend to move on a lag behind Saylor's own signals.

Our take

This isn't Strategy declaring victory. It's Strategy testing whether the market will still fund its bitcoin accumulation at a discount, which is a genuinely different question than the one it answered for most of the last two years. The company has never sold its full treasury and shows no sign of starting, but the June-through-August stretch proved the model has a real pressure point: preferred dividends need cash, and when bitcoin sold sideways instead of up, that cash had to come from somewhere. Buying back into the market this week says management thinks the worst of that pressure has passed. Whether that's right depends less on any Strategy decision and more on where bitcoin itself trades over the next month.

Primary sources

Original analysis by GenZTech, built from SEC filings and mNAV data cited above. Source: CryptoBriefing.