Bitcoin closed the week of 17 August around $63,548, down more than 3% over seven days. Ether sat near $1,907, off 1.8%. Against an October 2025 peak above $126,000, bitcoin is down roughly 50%.
Taken alone, that is unremarkable. Bitcoin has done this four times. Drawdowns of 50–80% are the asset’s normal behaviour, not evidence of failure, and anyone sizing positions off a different assumption was mispricing risk regardless of where the number lands.
What makes this cycle worth a closer look is not the depth of the drawdown. It is who is selling, and why the structure that was supposed to prevent that kind of selling is doing the opposite.
The Treasury Model Was a Leveraged Bet on Its Own Premium
The digital asset treasury company — a listed vehicle whose business is holding crypto — was the defining capital-markets story of the last cycle. The mechanism was straightforward and, while it worked, close to self-reinforcing:
- The stock trades at a premium to the value of the crypto it holds (mNAV above 1.0).
- The company issues equity into that premium.
- It buys more crypto with the proceeds.
- Crypto per share rises, which justifies the premium, which enables more issuance.
Galaxy Digital’s characterisation was that this was never a bitcoin strategy — it was a liquidity derivative that functioned only while the equity traded above the underlying. Once the premium compressed, the flywheel would run backwards.
It has run backwards more or less exactly as described.
Many treasury firms now trade below 1.0 mNAV — the company is worth less than the tokens on its balance sheet. Investors who bought the equity paid more for exposure than buying the asset directly would have cost. At Strategy, the largest of them:
| Metric | Reading |
|---|---|
| mNAV | ~0.63 |
| Enterprise mNAV | Crossed below 1.0 for the first time (~0.99) |
| Share price | ~$82, roughly 82% off peak |
| 2026 YTD | −43% |
| Q2 2026 | $8.2B book loss on bitcoin’s decline |
| Share of all DAT-held bitcoin | ~76% |
Enterprise mNAV falling below 1.0 is the reading that matters. It means total obligations — debt plus preferred stock, net of cash — now exceed the bitcoin. Above 1.0, the balance sheet is a bitcoin position with financing attached. Below it, the bitcoin is collateral against claims that are senior to shareholders.
And then the thing that was not supposed to happen: Strategy executed its first-ever bitcoin sale, reversing a “never sell” pledge that had been the entire identity of the strategy.
Why That Sale Matters More Than the Price
The bull case for treasury companies was that they converted volatile retail-held supply into permanent, price-insensitive corporate holdings. Structurally scarce float. Diamond hands with a ticker symbol.
The bear case was that they were leverage wearing a long-term-holder costume — and that leverage, wherever it sits, eventually becomes forced supply at the worst possible moment.
Strategy’s liquidation resolves that argument empirically. Preferred dividends and debt service are contractual. Bitcoin’s price is not. When those two facts collide and the equity window is shut because you are trading at 0.63 mNAV, the asset is what gives.
The concentration compounds it. With Strategy holding roughly 76% of all treasury-company bitcoin, this is not a diversified sector under pressure. It is one balance sheet, and its financing decisions are now a market-structure variable.
This is the same dynamic we flagged around the ETF outflow streak earlier this year, and it is why the record April inflows read differently in retrospect. Institutional demand did arrive. A meaningful share of it arrived through vehicles that had to keep buying to stay solvent — which is not the same thing as conviction, and behaves nothing like it on the way down.
What Is Actually Healthy
A drawdown is not a verdict. Several things are genuinely better than they were in the previous cycle, and it is worth being precise about which:
Failures have been institutional, not protocol-level. Nothing in 2026 resembles Terra or Celsius. No major stablecoin has depegged, no large lender has frozen withdrawals, and the chains have kept producing blocks through 50% drawdowns without incident. The infrastructure is doing its job.
Stablecoin rails have kept working. Post-GENIUS Act, the regulated stablecoin layer has functioned as payment infrastructure through the entire decline. That is the single most consequential product-market fit crypto has, and it is uncorrelated with price.
Smart-contract security genuinely improved. As we documented in the July recap, pure Solidity logic bugs are no longer where the money goes. A decade of tooling and audit practice worked. Attackers moved to firmware, cloud credentials, and social engineering — which is a real problem, but it is the problem of an ecosystem that fixed its previous one.
The leverage is unwinding in public. Strategy’s position is visible in SEC filings. Its mNAV is computable daily by anyone. Compare that to 2022, when the leverage sat in private balance sheets at Three Arrows and Genesis and nobody could see it until it detonated. Slow, legible deleveraging is a healthier failure mode than an opaque, sudden one — even though it feels worse for longer.
What Is Not
Price discovery is being distorted by a forced seller. When one entity holds three-quarters of a category’s supply and has contractual obligations to service, its liquidation schedule becomes a market variable rather than a company matter.
Retail participation has not recovered. Weekly leaderboards topped by staking-program launches and low-float tokens with triple-digit weekly gains are not a sign of health. They are what a market does when there is no directional conviction and speculation rotates into whatever is small enough to move.
Regulatory clarity did not arrive on schedule. Market structure legislation has slipped again — odds for the CLARITY Act have collapsed to around 10% for 2026 — and the institutional allocation that was supposed to follow the rules is still waiting.
The Read
Crypto’s asset prices are behaving normally for a bear market. Crypto’s capital structure is behaving badly, and it is doing so for an entirely predictable reason: the industry once again built a mechanism that only worked while prices rose, described it as structural demand, and is now discovering it was leverage.
The infrastructure is stronger than it was in 2022. The financialisation around it repeated 2022’s mistake in a form regulators approved and public markets funded.
The encouraging part is that this unwind is happening in filings rather than in a blowup. The discouraging part is that this cycle’s post-mortem will read almost identically to the last one, with different vocabulary.
Nothing here is investment advice.
Sources: Yahoo Finance — crypto prices 17 August 2026 · Yahoo Finance — Saylor’s treasury strategy · CoinDesk — DAT bitcoin buying · CoinGeek — digital asset treasuries · Bitcoin Foundation — Strategy Q2 loss
