Every Bitcoin Bear Market, Measured: How Deep and How Long

Everyone who holds bitcoin long enough lives through a crash. The question is what one is actually like.

Most charts show you how deep the crashes went. That's half the story. The other half is time: how long the fall lasted, and how long it took to get back. So we measured every fall of 20% or more from bitcoin's all-time high, using every daily close on Bitstamp from April 2013 to 30 September 2026.

Finding 1: by the stock-market rule, bitcoin has had 14 bear markets

In the stock market, a fall of 20% from the high is called a bear market. By that rule, bitcoin has had 14 since 2013.

But for bitcoin, 20% is normal. Eight of those 14 stopped short of 50%. All eight were back at a new high within 238 days. The quickest took eight days. Nobody in crypto calls those bear markets.

The ones people mean are the six that went past 50%. Four of them went past 70%. They followed the highs of April 2013, December 2013, December 2017, April 2021, November 2021 and October 2025.

Finding 2: the climb back takes longer than the fall

Look at the five finished falls of 50% or more. In four of them, getting back to the old high took longer than the fall itself.

After the 2017 top, bitcoin fell for 364 days. Then it took 716 more to get back. After the 2013 top, it fell for 406 days and climbed for 771.

So if you bought at the December 2013 top, you waited 1,177 days to break even. That's over three years. At the December 2017 top it was 1,080 days, and at the November 2021 top it was 847.

The exception was the April 2021 fall. Bitcoin fell 53% from its April 2021 high in 98 days, then got back in 91. That was a crash in the middle of a bull run, not the end of one.

Finding 3: half off wasn't the bottom

A 50% fall feels like it must be close to the bottom. In the three biggest falls, it wasn't even close.

In December 2017, bitcoin peaked at $19,188. By 1 February 2018 it had halved to $8,999. Then it fell another 65%, to $3,180, over 317 more days.

Here's why the last part of a crash hurts so much. Every percent you lose needs a bigger percent to win back. A 50% fall needs a 100% rise to get even. The 2021 fall of 77% needed 329%. The 2013 fall of 85% needed 560%.

Finding 4: bear markets have big rallies

A bear market isn't a straight line down. Inside the three biggest falls, bitcoin rallied 30% or more nine times before the bottom. Every one of those rallies gave way to a lower low.

The biggest one after the 2017 top ran from $6,874 to $11,463, up 67% in 27 days. In 2022, bitcoin rose 35% from $35,091 to $47,459, then fell to $15,766.

A rally like that feels like the end of the fall. In these three, it never was. The bounce looked like the turn, and the real bottom was still ahead.

Finding 5: most of the time, bitcoin is below its high

Since April 2013 there have been 4,931 daily closes. Only 171 of them set a new all-time high. That's 3.5%. Bitcoin closed within 10% of its high on 16% of days.

It closed at least half below its high on 43% of days. So if you own bitcoin, expect to spend a lot of your time looking at a big loss from the top.

Finding 6: where this one sits

Bitcoin's latest high close was $124,728, on 6 October 2025. Its lowest close since then was $58,526, on 30 June 2026. That's a 53% fall, the same depth as the April 2021 fall. On 30 September 2026 it closed at $83,563, still 33% below the peak, 359 days in.

That's a lot shallower than the big three at the same age. 359 days after their peaks, they were 67%, 82% and 70% down.

Each of the big three bottomed about a year after its peak, between 364 and 406 days. This one's lowest close so far came on day 267. It's also had one rally of 30% or more, from $62,795 in February to $82,193 in May, before the June low.

None of this tells you whether the low is in. Four big falls is too few to set a timetable. What it does tell you is the size of the job: to get back above $124,728, bitcoin needs to rise 49% from here.

Our study of the halving cycle found that each cycle's crash has been smaller than the one before. This fall fits that pattern so far.

How to read it

• Measure time, not just depth. The question isn't only how far it falls, it's how many years you could be underwater.
• Don't treat half off as a floor. In the three biggest falls, a 50% drop came long before the bottom.
• Expect big rallies on the way down. A 30% bounce inside a bear market is normal and tells you little.
• Do the recovery maths. A 75% fall needs a 300% rise to get back, so size your position for that.
• If you buy a fixed amount every month, a long fall means many of your buys land at low prices. That's the case for dollar-cost averaging into something this volatile.

Method, and what this cannot tell you

We took every daily close of bitcoin on Bitstamp, in US dollars, from September 2011 to 30 September 2026. A fall starts at a record close and ends at the first close back above it. Its depth is the lowest close in between. We count falls from April 2013, because Bitstamp's record starts after bitcoin's first big peak in June 2011, and by April 2013 the price was far above that peak. Binance's daily closes start in 2017, and they give the same peaks and lows for every fall of 50% or more since then.

• Closes only, in UTC. Intraday lows went deeper than any close here. A depth measured from the highest and lowest trade would be bigger.
• Four falls of 70% or more is a small sample. The next one doesn't have to look like any of them.
• Bitstamp was a small exchange in 2013, and prices from those years differ from one exchange to the next.
• Falls before April 2013 are left out, including bitcoin's first big crash in 2011, because our record starts after its peak.
• The 2025-26 fall wasn't finished on the study date. Its numbers are "so far", and its lowest close may not be its last.
• A 30% rally is our cut-off. A different cut-off would count a different number of rallies.