Bitcoin is down about half from its high. Is the 2026 bull run over?

If you only saw the headlines, you would think crypto had collapsed. Bitcoin has lost about half its value, altcoins have fallen harder, and more than two trillion dollars has vanished from the market since last autumn. That is real, and if you bought near the top it hurts. But panic and context are different things, and the honest version of this story needs both. Here is how far Bitcoin has actually fallen, what pushed it down, and why serious people disagree about whether the bull run is over. None of this is advice about what to do with your money; it is an explanation of what happened and what is being argued.

How far has Bitcoin fallen?

A lot, but the number that matters is the comparison, not the raw drop.

Bitcoin peaked at about $126,000 in October 2025 and now sits around $63,800, a fall of roughly 50% (estimates range from about 48% to 52% depending on the day). Across the whole market the damage is larger: the total value of all crypto fell from a peak near $4.3 trillion to roughly $2.3 trillion, erasing well over $2 trillion.

Here is the twist the doom headlines skip. In Bitcoin's short, violent history, a 50% drawdown is mild. The 2021-22 bear market cut it 77%; the 2017-18 crash, 84%; earlier ones, more than 90%. Measured against its own past, this is the gentlest major sell-off Bitcoin has ever had. That does not make it pleasant, but it reframes "collapse" into something closer to "a rough but ordinary crypto winter."

What actually caused the crash?

There is no single villain. The fall is the sum of several macro shocks landing on an asset that had run up hard, and it is worth separating them because each is real:

• Tariffs. In February 2026, a new round of US tariffs (raised to a 15% rate) hit risk assets, and Bitcoin fell sharply on the day as markets went risk-off.
• The Fed stopped cutting. With inflation running in the high 3s to low 4s (around 3.8% to 4.2%), traders spent 2026 pricing in essentially no Federal Reserve rate cuts. Higher-for-longer rates are poison for speculative assets, crypto included.
• The AI stock sell-off. This is arguably the most important. Bitcoin now trades almost in lockstep with US stocks (its correlation with the S&P 500 hit a record of around 0.96 in early 2026), so when AI-driven chip stocks sold off, Bitcoin went with them. The same AI-bubble anxiety shaking equities is shaking crypto.
• Geopolitics. Repeated US-Iran flare-ups over the Strait of Hormuz through mid-2026 triggered sharp risk-off moves and hundreds of millions in liquidations on the worst days.
• Forced selling. A cascade of leveraged liquidations, including a roughly $19 billion wipeout back in October 2025 that analysts say damaged market structure, plus multi-billion-dollar liquidation days in 2026, repeatedly knocked the price lower as over-leveraged bets were force-closed.

The through-line: in 2026 Bitcoin behaved less like "digital gold" and more like a high-risk tech stock, and it fell for the same reasons tech stocks did.

Are the big institutions bailing out?

This is the swing factor, and the picture is mixed rather than a clean exodus.

The spot Bitcoin ETFs that drove the 2024-25 rally cooled hard through the summer. July 2026 saw the weakest monthly inflows on record (barely positive), after net outflows of about $2.4 billion in May and $4.5 billion in June. Then, in early August, demand snapped back: spot Bitcoin ETFs took in roughly $853 million in the first week of the month, their biggest weekly inflow since April, with BlackRock's IBIT alone drawing about $693 million. Long-term holders have been selling into the weakness too, so the flows have been volatile in both directions rather than a one-way exodus.

That volatility is the point: "cooling" is not "leaving." The ETF rails, custody infrastructure and clearer US rules that brought big money in still exist, and August's rebound shows the demand can return fast. Whether institutions are taking profits before re-entering, or heading for the exit, is exactly what the bulls and bears are fighting over.

Is the bull run actually over?

Honestly, nobody knows, and anyone who tells you they do is guessing. Here is the real disagreement.

The bears see a cycle topping into a genuinely hostile backdrop: no rate cuts, sticky inflation, record ETF outflows earlier in the year, structural damage from the October liquidation cascade, and a Bitcoin that now falls every time AI stocks do. In that reading, the easy money is gone and this grinds lower.

The bulls see a normal, even shallow, correction. Bitcoin has always had 30-50% pullbacks inside its bull markets, this drawdown is milder than any previous big one, and the four-year halving cycle that has front-run every prior peak is intact; the April 2024 halving preceded this top almost exactly on schedule, which by past patterns would put a bottom in late 2026 or early 2027. Firms like 21Shares openly frame the current levels as a possible cycle bottom rather than a top.

Both cases are built from real facts. That is why "is the bull run over" has no honest yes-or-no answer right now, only a probability that reasonable people weigh differently.

Bitcoin's crashes, in context

Bear market	Peak-to-trough drop
2011	about -93%
2013-2015	about -85%
2017-2018	about -84%
2021-2022	about -77%
2025-2026 (so far)	about -50%

The honest bottom line

Crypto in 2026 is having a bad year, not obviously a fatal one. Bitcoin is down about half, the market has shed over $2 trillion, and the causes are real and mostly macro: tariffs, a tighter Fed, an AI-stock sell-off and geopolitics, amplified by leverage. At the same time, this is the mildest major crash in Bitcoin's history and the long cycle it has followed for a decade has not broken. The useful stance is neither "it's over" nor "buy the dip," but the boring truth in between: this is a real drawdown with real macro causes, and whether it is a bottom or a top will only be obvious in hindsight. For more, see the Crypto section, our guide to buying and holding crypto safely, and the related question of whether AI itself is a bubble.