If you’re reading this in the summer of 2026, you’re probably sitting in the middle of it: Bitcoin has traded below its October high for a good nine months, at times down by more than half. And anyone looking for orientation in this situation meets two kinds of texts: pep talks (“cycles repeat!”) and obituaries (“this time everything is different!”). Both claim to know the future.

This is the third way: simply looking up what actually happened. Every completed Bitcoin bear market, with dates, depth and duration — and afterwards the uncomfortable question of what four data points actually prove. (Spoiler: less than you’d like them to in a drawdown.)

Two definitions up front, so the counting doesn’t look arbitrary: I count cycles from all-time high to all-time high — not every 20% correction, or the list would be three times as long. By that convention there have been four completed bears. The 2019/20 slump (from an interim high of ~$13,800 into the COVID crash at ~$3,900) is deliberately absent from this count: it happened below the old 2017 all-time high — by the high-to-high convention it’s a chapter of the long 2017 cycle, not a cycle of its own. That’s a convention, not a truth; I’m just saying it out loud so you can check it. Also: all prices in this post are US dollars — in euros the percentages look somewhat different due to exchange rates.

A vast frozen sea under a deep grey winter sky, broken ice sheets stretching to the horizon, a tiny lone figure walking across the ice

Illustrative image, AI-generated.

The four completed bears

Behind every row sits its own story — and those are worth telling, because they show that “the cycle” was never the same thing twice.

2011 — the first crash

From the June high of around $32 — measured on Mt. Gox, the then-dominant exchange — down to around $2 by November. Minus 93 percent, to this day the deepest bear in history. One famous event needs to be cleanly separated from the bear market itself: the “flash crash” to $0.01 on June 19, 2011. That was an exchange-internal hack event on Mt. Gox, dissected by BitMEX Research, and the trades were rolled back — the real bear played out over months, not minutes. Remarkable in hindsight: a new all-time high stood only about 21 months after the peak — the fastest reconquest of any cycle.

2013–2015 — the long desert

From the high of around $1,150 in late November 2013 down to $152 on January 14, 2015 (Bitstamp; other exchanges somewhat higher). Shaped by the collapse of the same exchange that delivered the 2011 flash crash: Mt. Gox shut down in February 2014, with some 850,000 customer and company BTC presumed lost. Above all though: 39 months from peak to new all-time high in early 2017. If nine months of drawdown feel long — this was the benchmark for long.

2017–2018 — the ICO bubble bursting

From just under $20,000 in December 2017 to around $3,100 almost exactly one year later, minus 84 percent. The bear that presented the altcoin bill: Ethereum fell about 94 percent over the same period, and nearly half of that era’s top 10 — Bitcoin Cash, NEM, Stellar, IOTA — never saw another all-time high even in the next bull market. A documented long-term experiment that put $100 into each of the January 2018 top 10 had, after six and a half years, only Bitcoin and Ethereum in the green. XRP needed about seven years to revisit the level of its January 2018 high. That’s the part that backward-looking “crypto always recovers” charts systematically omit: they only show the survivors.

2021–2022 — the cascade

From the high of around $69,000 in November 2021 to around $15,500 a year later, minus 77 percent — the “mildest” bear on the list, and at the same time the one with the most brutal mechanics. It came in waves: in May 2022, Terra/Luna collapsed — over 40 billion dollars of market value erased within a week. The insolvency wave ate through entangled, opaque credit: Celsius froze withdrawals in June, the hedge fund Three Arrows was liquidated. And in November, a single CoinDesk investigation into Alameda’s balance sheet brought FTX down within nine days — only after that, on November 21, did the cycle bottom stand. New all-time high: March 2024, after 28 months.

The current one — without a forecast

And now the cycle still running — with due caution: it is not completed and does not belong in the statistics above — its row in the table is deliberately marked open.

What’s verifiable: an all-time high on October 6, 2025 at around $126,200. A first capitulation wave in early February 2026 down to just above $60,000. The trough so far on June 25, 2026 at around $58,100 — nearly 54 percent below the high. As of July 23, 2026, Bitcoin trades about 48 percent below the October high; for today’s price, check live — on a static page it would be wrong by tomorrow.

Whether the June trough was the trough, nobody knows — including the people claiming it. The four historical bears took between 5.5 and 13.5 months from peak to trough; the current one is at month nine. That’s a reading of where we stand, not a statement of where it goes.

A young man in a light tee sits calmly at his desk before two screens showing falling red price charts, warm lamp light

The survivor tee from the Crypto Collection — holding on, without promises. Illustrative image, AI-generated.

What measurably happens in bears

Beyond prices, there are patterns observable on-chain — as description, mind you, not as a signal:

A woman in a black cap with a small white bear silhouette looks calmly out the window of a concrete room

The bear cap from the Crypto Collection. Illustrative image, AI-generated.

The small ones stack when it hurts

In the 2022 bear, the smallest holder classes absorbed more coins than were even newly mined, according to Glassnode — at record pace right after the FTX collapse. If you want to meet the sea-creature taxonomy behind that: Shrimp, Fish, Whale — who actually holds the coins. The pattern has so far repeated in the current drawdown; accumulation scores of the smallest classes stood near maximum in early July 2026, per CoinDesk/Glassnode.

Long-term holders grow, not shrink. The share of coins held by long-term holders — by Glassnode’s 155-day heuristic, a convention, not a law of nature — rose in each of the 2015, 2019 and 2022 bears. It’s the short-term holders who sell; and in 2022 the pain realized along the way was historic: the June crash produced what Glassnode called the largest daily realized loss in Bitcoin’s history — around 4.2 billion dollars in a single day.

The machines capitulate too. In every hard bear, unprofitable miners switch off; in 2022, listed miners went bankrupt. (The most famous hashrate collapse — minus 50 percent in the summer of 2021 — does not belong in this series, by the way: that was China’s mining ban, a regulatory shock, not a price event. The hashrate stood at a new all-time high by December 2021.)

And the obituaries arrive on schedule. The half-ironic “Bitcoin Obituaries” counter at 99Bitcoins lists, as of July 23, 2026, 477 media death notices since 2010 — their density has always peaked in bear markets. It’s a curated collection with a wink, not a dataset. But as an archive of sentiment, it’s priceless.

Why all of this proves less than it seems

Now for the part that’s usually missing from cycle threads.

Four is almost nothing. Four completed bears, each with a different trigger (exchange hack, exchange collapse, burst ICO bubble, credit cascade), different market size, different environment. You cannot derive laws from n=4 — every “cycle rule” is pattern-matching on a sample that would be called anecdotal in any other discipline.

Survival proves nothing. That Bitcoin has come back from every bear so far is a historical fact. What follows from that for the future: logically, nothing. That’s the problem of induction, and it doesn’t get smaller by looking good on a poster. Add the survivorship bias at the asset level: the altcoin graveyards of 2018 and 2022 show what “crypto always recovers” looks like once you also count the ones that never did.

And this cycle is structurally different. Since January 2024, spot Bitcoin ETFs trade in the US — for the first time, a regulated capital-market channel hangs directly on the price. Whether that makes future bears milder, harsher, or just different: open. The debate over whether the four-year cycle still exists at all, or lives on only as a self-fulfilling prophecy, is being conducted openly among analysts and is not settled. I count myself among those who simply don’t know.

So what remains of this post? This: drawdowns of 77 to 93 percent and dry spells of 21 to 39 months are not an accident in this market — they are its documented normal. Knowing that beforehand leads to different decisions than finding it out in month nine; which decisions, remains your business. What is concretely being built during the current bear is documented in the market report — as a report, not as consolation. And honestly enduring that uncertainty, without promises: that is exactly what the survivor motif of the Crypto Collection stands for — no promise, just a stance.

A framed poster showing an infinity sign above the number 21,000,000 hangs on a concrete wall above a wooden sideboard, a potted plant beside it

The “Scarcity” poster (the 21-million cap) from the shop — the stance on the wall. Illustrative image, AI-generated.

Sources and limits

The cycle data rests, where possible, on contemporary reporting and primary research (BitMEX Research, CoinDesk — including the Alameda investigation that triggered the FTX collapse —, NPR, CNBC, TechCrunch, Forbes, Glassnode). Historical intraday extremes differ by exchange — in 2013, Mt. Gox traded at a marked premium — which is why this post uses rounded values and ranges throughout; if you need cent amounts, you need an exchange and index attached. All month figures are computed from the documented dates themselves. The current cycle is marked open in both table and text; its “trough” is the low so far, nothing more. And once again: nothing here implies an action — not buying, not selling, not holding.

Questions, or spotted a mistake? Write to me.