The short version
- Missing bitcoin's 10 best trading days per year historically turns a ~90% median gain into a ~25% loss, per analyst David Eng's 2020–2025 study cited by CoinDesk.
- In 2026, removing just 5 best sessions deepened bitcoin's year-to-date loss from roughly 9% to roughly 36%.
- In 2019, buy-and-hold returned roughly 94%; missing the 10 best days would have produced a loss of roughly 40%.
- Historical back-tests show holding bitcoin for 3+ years reduced the loss probability to below 1% — a back-test result, not a forward-looking guarantee.
How Bitcoin stacks its annual gains in days, not months
An analysis of Bitcoin's daily returns from 2020 through 2025 found that the median annual return for an investor who held through every session was roughly 90%. Remove the ten best-performing days from each year, and that same investor would have finished with a median loss of around 25%. The study, attributed to analyst David Eng, was cited in a CoinDesk report published September 5, 2026.
Andre Dragosch, Head of Research at Bitwise Europe, put the pattern plainly: most of the time, bitcoin moves sideways and consolidates, then delivers its performance in a handful of sessions. In a typical year, the gap between a good outcome and a bad one can come down to whether an investor happened to be in the market on a few specific days.
The same underlying pattern exists in equity markets. JPMorgan Asset Management has published research showing that missing the ten best trading days on the S&P 500 over a twenty-year period roughly halved terminal returns for equity investors. Bitcoin's version of this effect is substantially more severe. Equity markets spread gains across more sessions; bitcoin delivers a larger fraction of its annual return in far fewer bursts.
Check Bitcoin’s current reference price
A year in bitcoin can swing on a single weekend
Adam Haeems, Head of Asset Management at Tesseract Group, offered a concrete example from February 2026. On February 5, bitcoin fell roughly 14% in a single day. The following day, February 6, it recovered roughly 12%. An investor who sold after the first day's drop — a natural reaction — would have missed the full recovery and locked in a loss that the market erased within 24 hours.
That sequence reflects a wider pattern across Bitcoin's history. Early exchange data from 2011 is not available from a primary source for independent verification, but figures cited in the CoinDesk report indicate that year's buy-and-hold return of roughly 1,474% shrank to about 2.2% without the ten best days. Nearly the entire gain was concentrated in fewer than two weeks of trading across a full calendar year.
The table below draws on figures cited consistently across the CoinDesk report and multiple secondary outlets. Each row shows how dramatically the outcome shifts once the best sessions are excluded. In 2026, bitcoin was tracking a loss of roughly 9% through early September on a buy-and-hold basis; strip out just five of those sessions and the loss deepens to roughly 36% — a gap of 27 percentage points.
| Period | Sessions removed | Return without them |
|---|---|---|
| 2019 | 10 best days | −40% |
| 2020–2025 median | 10 best days/year | −25% |
| 2026 YTD (early Sep.) | 5 best days | −36% |
Learn how Bitcoin’s market price is formed
Why Bitcoin's volatility concentrates returns in so few sessions
Bitcoin trades around the clock, every day of the year, on exchanges worldwide. Unlike stock markets that close on weekends and holidays, bitcoin can move sharply at any hour. The asset's price volatility is substantially higher than most conventional investments, and that volatility does not spread evenly across sessions. A few days each year account for movements that stock markets might take months to deliver.
News events, large institutional trades, liquidations of leveraged positions, and shifting market conditions can all trigger outsized single-day moves. When highly leveraged traders are automatically forced to sell, prices can move far and fast, and a sharp reversal often follows just as quickly. Both the drop and the recovery tend to happen in compressed timeframes, concentrating gains into just a few sessions per year.
Bitcoin's daily return distribution has what statisticians call 'fat tails.' Extreme outcomes — very large single-day gains or losses — happen far more often than a normal distribution would predict. The best days are not just slightly above average; they are genuinely exceptional. That is why the gap between a buy-and-hold strategy and one that misses just ten sessions can exceed 100 percentage points in a single year.
- Forced liquidations of leveraged positions, which can trigger cascading price moves and equally rapid reversals
- Macro news and policy announcements that arrive outside stock-market trading hours, when bitcoin keeps trading
- Large institutional orders that move thin order books quickly, pushing prices far from recent levels
- Short squeezes that drive sharp recoveries after a sustained period of selling pressure
Compare the wider Bitcoin and crypto market
2013 and 2017: when Bitcoin rallied broadly enough to change the math
Not every year follows the same pattern. The analysis identified 2013 and 2017 as exceptions: in both years, Bitcoin's gains spread broadly enough across the calendar that removing the twenty best trading days still left investors with a positive annual return. That is unusual. In most years covered by the data, removing just the ten best days is enough to flip a positive year into a losing one.
Both 2013 and 2017 were years of sustained, multi-month rallies. In 2013, bitcoin rose broadly across the full calendar year, with demand building at several points across different months. In 2017, a broad rally carried the asset from below $1,000 in January to near $20,000 by December, with gains distributed across many sessions rather than clustered in a handful of extreme outliers.
That distinction matters: when a demand shift plays out over months, gains distribute broadly and missing one session costs less. But investors cannot know in advance whether a given year will resemble 2017 — with broadly spread gains — or 2019, where the ten best days delivered nearly the entire 94% annual return. The DCA backtest tool lets users explore how different entry points have performed historically.
Convert a Bitcoin amount using a reference rate
Why those ten days make timing the market so difficult
A historical back-test found that buying bitcoin and holding for three or more years reduced the probability of finishing with a loss to below 1%, based on all historical entry points in the data. That is a back-test result, not a forecast — it ignores scenarios outside the historical record. Even so, it helps explain why many holders who follow the bitcoin price daily still choose not to trade in and out.
Timing the market requires getting two decisions right: when to sell, and when to buy back in. An investor who sells during a drawdown then needs to re-enter on the best days — which cannot be predicted. The February 2026 example illustrates this directly: the session that gained roughly 12% came the day after a roughly 14% drop. Selling on day one meant missing the recovery entirely.
Andre Dragosch's description of bitcoin as 'a relatively boring asset' that moves sideways most of the time explains why the pattern persists. If most days produce modest moves, and a small number of sessions deliver extreme gains, staying invested through quiet periods is how long-term returns are built. Missing ten days out of roughly 365 sounds minor. The 2020–2025 data shows it is not.