Half of all bitcoin is now underwater
1. What happened: no crash, just a nine-month bleed
There is no single blow-up day to point at. After touching $126,080 in October 2025, bitcoin went sideways, then stepped down, month after month. What put the market on alert was three technical events stacking up: on July 1 the price broke $58,000 intraday, a 21-month low; the weekly candle closed below the 200-week moving average for the first time in four years — the line that acted as the bear-market floor in the past two cycles; and since June 5, Glassnode's measure of "supply held at a loss" has exceeded "supply held in profit."
That last one deserves unpacking. Supply-in-loss treats the price at which each coin last moved on-chain as its holder's cost basis, then compares it with the current price. When the crossover happened in early June, the share of supply at a loss had jumped from 34% to over 50% in a single month. By early July, roughly 10.83 million BTC were underwater against 9.22 million in profit. In plain terms: the majority of coins in circulation are held by people who are losing money on paper. Since 2010, this condition has appeared only in deep bear markets — 2011, 2014–15, 2018, March 2020, and 2022.
2. Why: four pipes reversed at once
First, macro. The new Fed chair has taken a plainly hawkish line, inflation is stuck above target, and markets have abandoned hopes of cuts this year — rate swaps have started pricing hikes instead. A risk-free rate above 3.5% is a constant drain on an asset that yields nothing. The 2022 playbook is running again, at a higher level of rates.
Second, the ETFs — this cycle's genuinely new variable. When spot ETFs were approved in early 2024, the prevailing story was "a one-way pipe for traditional money." Pipes flow both ways. U.S. spot bitcoin ETFs bled $4.06 billion in June, their worst calendar month since launch, including a 10-day streak of uninterrupted outflows. The ETF wrapper lets retirement accounts and advisor platforms buy with one click — and redeem with one click, with far more mechanical stop-loss discipline than crypto-native holders ever had. For the first time, the selling has a fully compliant channel.
Third, the faith anchors loosened. In June, Strategy (formerly MicroStrategy) disclosed its first bitcoin sale since December 2022 — just 32 BTC, about $2.5 million, trivial in size but heavy in symbolism, because the company's role for four years was to be the permanent bid. The premium at which bitcoin-treasury companies trade above their coin holdings (mNAV) deflated fast. The same month, Mt. Gox moved about 10,422 old coins (~$739 million) to a new wallet, reviving creditor-selling fears ahead of the October repayment deadline and triggering $1.86 billion in liquidations in a day.
Fourth, opportunity cost. Over the same six months, the Nasdaq rose 12% and chip stocks printed their best quarter ever, while bitcoin lost a third of its value. Money didn't flee risk — it moved to AI. That's a double failure for bitcoin's two competing stories: it didn't rally as "digital gold" while actual gold climbed to $4,190, and it didn't track the Nasdaq as a high-beta risk asset either. An asset that tells two stories and delivers on neither gets a valuation discount for the confusion itself.
3. Second-order effects: who's hurt, who benefits
- Hurt most are this cycle's newcomers: traditional investors who entered high through ETFs in 2024–25 — the people just persuaded by their advisor to "allocate 2%," with cost bases likely above $80,000; post-halving miners whose margins compress dangerously below $60,000; and Strategy imitators whose equity premium over their coin holdings has flipped to a discount.
- Benefiting is patient old money: the largest wallets absorbed roughly 270,000 BTC (~$16.7 billion) in the last two weeks of June. Panic supply is migrating from levered and new hands to holders indifferent to volatility — the handover that happens every bear market, except this time the sellers include a brand-new seat called the ETF.
- For ordinary savers, the direct lesson is that 2024's popular claim — "institutionalization makes it safer" — has been disproved. Institutional money didn't reduce the volatility; it changed how the volatility travels: from cascading exchange liquidations to a quiet daily bleed in ETF flow tables.
4. The strongest case on each side
Bulls: the 50/50 loss crossover has appeared four or five times since 2010 and each instance mapped to a bear-market bottom zone; deeply underwater holders historically sit tight rather than sell, so marginal sell pressure dries up; whale accumulation of 270K BTC is real money; and on July 3, ETFs took in $221.7 million — the largest daily inflow in two months, snapping the 10-day outflow streak, with Fidelity's FBTC alone drawing $166 million.
Bears: the same "bottom signal" persisted for nearly a year in 2015 and about six months in 2022 before price actually bottomed — it marks a zone, not a date; bitcoin hasn't yet broken below its $54,000 realized price, and every major historical bottom formed below it, implying the capitulation flush may still be ahead; Glassnode itself calls this "the early stages of a bottoming process," noting past cycle lows usually ended with one final volatility spike; and there are no rate cuts coming to supply the liquidity for a rebound.
5. What the narrative glosses over
The most popular line right now is "every time supply-in-loss crossed 50%, it was a buying opportunity." Two problems. First, the sample size is four or five, and every instance was labeled in hindsight — the same family of indicators flashed for months through late 2022 while price fell another 30%. Extrapolating from five observations is storytelling with extra steps. Second, this is the first out-of-sample cycle: every previous crossover happened in a world with no spot ETFs and near-zero rates. This time the sellers have a compliant institutional exit, and the opportunity cost is a 3.5% Treasury and an AI equity market in full sprint. History rhymes — but only when the rhyme scheme survives.
The bear narrative has its own inflation, though: "losing the four-year floor means the cycle is dead" treats a chart level as a cause. The holder structure, the whale accumulation, and the flip back to ETF inflows all suggest a violent change of hands rather than an asset being abandoned.
6. One lesson worth keeping
For any asset, don't just ask "how far is it below the high?" Ask "who holds the supply, and at what average cost?" For bitcoin today the answers are: half the coins are underwater, and the network's average cost basis is $54,000. Those two numbers are more honest than any analyst's price target. And for your own portfolio, the only reliable stress test is: a position is only truly yours if a 50% drawdown doesn't cost you sleep. Narratives are loudest at tops and silent at bottoms; what carries you across both is position sizing, not conviction.
Sources
- CoinDesk — Bitcoin supply in loss overtakes profit, a hallmark of bear-market bottoms (2026-06-04)
- Glassnode — The Week On-chain #26 2026: Accumulation Beneath the Surface
- CoinDesk — $221M flows into Bitcoin ETFs, ending a 10-day outflow streak (2026-07-03)
- CoinDesk — Live updates: more bitcoin is now held at a loss than at a profit (2026-07-03)
- Investing.com — Bitcoin falls as record ETF outflows and Strategy sale hit sentiment
- Bitcoin.com News — Bitcoin falls below $58,000 as Q2 2026 losses deepen