What If You Bought Bitcoin at the All-Time High?

The interesting question about a peak is never what you lost. It is what you have to believe to keep holding. Bitcoin set its all-time high of $126,080 on October 6, 2025. As of July 8, 2026 it last traded at $61,678, roughly 51.1% below that peak. Nine months, about half the money.

Someone bought at the top. Someone always does, and it is rarely because they were reckless. The top is the moment the story is most convincing. So here is the arithmetic, run straight, along with the times Bitcoin has climbed out of a hole like this before and what that does and does not tell you.

What a peak purchase is worth

The math has no trick in it. At $126,080 per coin, $10,000 buys 0.07931 BTC. Multiply 0.07931 by the July 8, 2026 price of $61,678 and you get $4,892. The coin count never changed. Only the price did.

The hole is deeper than the percentage sounds

A 51.1% decline does not need a 51.1% rise to undo it. It needs +104.4%. Bitcoin has to slightly more than double from $61,678 just to return the peak buyer to even, before any gain at all. That asymmetry is the whole reason drawdowns matter more than the headline number suggests, and it is why the size of a loss and the work required to erase it are two different facts.

The times Bitcoin has done it before

Bitcoin has set a cycle peak, collapsed, and eventually reclaimed that price four times. CoinGecko's cycle research dates three of them: a November 2013 peak of $1,127, reclaimed on January 4, 2017; a December 2017 peak of $19,665, reclaimed on December 1, 2020; and a November 2021 peak of $69,044, reclaimed on March 8, 2024. The fourth is older and rougher. Bitcoin peaked at $32 in June 2011, and Forbes records a long bear market that bottomed at $2 in November 2011, a total decline of 94%, with prices not returning to that peak until 2013.

The three most recent are worth stating precisely, because the stories people tell about them tend to round the timelines down.

Peak prices and reclaim dates for the first three cycles: CoinGecko's cycle research, which reads daily closing prices, so both ends of each span sit on one basis. The October 6, 2025 figure is CoinGecko's intraday all-time high, a different and slightly higher basis. Trough column: the lowest completed month-end close in this site's Yahoo Finance monthly series, which begins in September 2014, ten months after the 2013 peak.

The 2013 peak. Bitcoin topped out at $1,127 on November 30, 2013. This site's monthly series picks the story up in September 2014 and records a low close of $217 in January 2015, about 80.7% below the high. The peak was not reclaimed until January 4, 2017, roughly 37 months later.

The 2017 peak. Bitcoin reached $19,665 on December 16, 2017, then fell to a low month-end close of $3,458 in January 2019, a drop of about 82.4% from the high. The peak was reclaimed on December 1, 2020, and the round $20,000 broke on December 16, 2020. Call it just under three years.

The 2021 peak. Bitcoin reached $69,044 on November 10, 2021, bottomed at a $16,548 month-end close in December 2022, about 76.0% down, and reclaimed the peak on March 8, 2024. That took roughly 28 months, about two years and four months.

Every one of those recoveries was real, and every one was slower and uglier than the retelling. A peak buyer in December 2017 spent thirteen months watching the position lose more than four fifths of its value before anything went right. Nine months into the current drawdown, the current peak buyer is earlier in that story than it feels.

Four recoveries is not a pattern

This is the part the recovery charts leave out. Bitcoin has completed four peak-to-recovery cycles. Four. That is not a sample you can build a probability from, and nothing about the previous four obliges a fifth to happen, or to happen on a similar schedule. Each of those recoveries also played out under conditions that were specific to its moment, and an asset can recover four times and still fail to recover the fifth. Past recoveries do not guarantee a future one.

What the record does support is narrower and duller: Bitcoin has been this far below a high before and did not stay there. That is a statement about what has happened, not a forecast of what will. Anyone telling you the next recovery is due is reading four data points as destiny.

What spreading the money out would have changed

The peak buyer put the whole stake in on one day. Consider the other version: the same $10,000, split evenly across every completed month-end close from October 2025 through June 2026, about $1,111 a month. On this site's month-end data that buyer ends up holding 0.13077 BTC at an average cost of $76,471 per coin, worth $7,658 as of the June 2026 close, or -23.4% on the full $10,000.

For comparison, $10,000 dropped in at the October 2025 month-end close of $109,556 would be $5,345, or -46.5%, measured to that same June 2026 close. Spreading it out cut the loss roughly in half: 23.4% instead of 46.5%.

Read that honestly. Averaging in did not beat the lump sum because averaging in is clever. It won because the price fell, and buying a falling price at nine different levels beats buying it once at the top. Run the same exercise across a period when Bitcoin rose and the lump sum wins by the same mechanism, which is what the S&P 500 version of this question usually shows. Averaging in is a way of being less wrong about timing, in both directions. It is not a way of being right.

And the version of the question that actually faces a reader today, whether to start averaging in from here, is one no dataset can answer. Every figure above describes a price path that already happened. The cost basis you would build starting this month depends entirely on a price path that has not.

Meanwhile, the boring comparison

Over the eight months from the October 2025 close to the June 2026 close, $10,000 in the S&P 500 (via SPY) would be $11,011, a +10.1% return with dividends reinvested. The same $10,000 in Bitcoin, on that same month-end basis, would be $5,345, down 46.5%. That is not an argument that the index is better. It is a reminder that the stretch which took nearly half of one asset was unremarkable for another.

How these numbers were built

Two bases, kept separate on purpose. An all-time high is an intraday print, so the peak and the last-traded price come from CoinGecko, read on July 8, 2026: an all-time high of $126,080 on October 6, 2025 and a last trade of $61,678. Everything on a month-end basis (the averaging-in figures, the cycle lows, the SPY line) comes from this site's Yahoo Finance monthly closes, complete through June 2026, where Bitcoin closed at $58,559. The two differ because one is a July 8, 2026 tick and the other is June 2026's closing price. The SPY series is dividend-adjusted, which is why its return is a total return; Bitcoin pays nothing, so its figures are price-only either way.

Exact tops disagree across sources. Rather than quietly picking the flattering one: CoinGecko's coin page records the October 6, 2025 high at $126,080, an intraday print, while CoinGecko's own cycle research, which reads daily closing prices, dates the latest all-time high at $124,128 on August 14, 2025. The same split shows up in 2017: that research dates the top at $19,665 on December 16, while the CoinDesk Bitcoin Price Index recorded an all-time high of $19,783.21 on December 17. A gap that small, about 0.6%, is enough to move a reclaim date by days, which is why the cycle table takes each completed peak and its reclaim from the same source rather than pairing the highest peak anyone reported with the earliest reclaim anyone reported. A monthly close series cannot settle those dates at all. On this site's month-end data, Bitcoin's first monthly close above the 2017 top came in December 2020, after a November 2020 close of $19,626 that landed just under it.

The bottom line

A $10,000 Bitcoin purchase at the October 6, 2025 all-time high of $126,080 was worth about $4,892 on July 8, 2026, down 51.1%, and getting back to even requires +104.4% from there. Bitcoin has climbed out of four holes deeper than this one. The three the price record dates cleanly took about 37 months, about 35 months, and about 28 months. Whether it climbs out of a fifth is not something the first four can tell you. This is historical math, not financial advice, and past performance does not guarantee future results.

For the longer view, see what buying Bitcoin at $10,000 would be worth, how Bitcoin stacks up against the S&P 500 over a full holding period, and what happened to people who bought at the worst possible moments in stock market history. You can run any entry month in the calculator.