Crypto seasonality: the best and worst months
"September is bad, Q4 is good" is repeated every year. We measured it: the average return of each calendar month for Bitcoin over 15 years and for the 10 largest cryptocurrencies, from real daily closes.
avg +36.5% Β· green 60% of years
avg β1.4% Β· green 43% of years
Key findings
- Bitcoin's strongest calendar month over 15 years is November (avg +36.5%); its weakest is September (β1.4%).
- The fourth quarter carries the year: October averages +14.4% for Bitcoin, and November +36.5%.
- The mid-year is the soft patch β September is green only 43% of years for Bitcoin, and across the 10 majors the weakest month is June (β8.6%, green just 21% of the time).
- It is not only a Bitcoin quirk: pooled across the 10 largest coins, the strongest month is January (+22.9%) and the weakest is June.
- Averages are skewed up by crypto's early explosive years, which is why the positive-rate β how often a month was actually green β matters as much as the mean.
Every month, measured
Average month-over-month return and how often the month closed green, for Bitcoin and for the 10 largest coins pooled together.
| Month | BTC avg | BTC green | Majors avg | Majors green |
|---|---|---|---|---|
| January | +4.4% | 53% | +22.9% | 54% |
| February | +10.0% | 67% | +14.9% | 54% |
| March | +11.6% | 47% | +6.5% | 54% |
| April | +11.2% | 67% | +15.2% | 58% |
| May | +8.9% | 53% | +3.2% | 48% |
| June | β0.1% | 53% | β8.6% | 21% |
| July | +9.8% | 73% | +10.9% | 67% |
| August | +3.1% | 40% | +4.6% | 43% |
| September | β1.4% | 43% | β0.8% | 44% |
| October | +14.4% | 67% | +8.7% | 60% |
| November | +36.5% | 60% | +19.7% | 51% |
| December | +6.9% | 47% | β1.7% | 37% |
"Green" is the share of years that month closed higher than the previous month-end. Bitcoin spans 15 years; the majors pool covers a shorter history for the newer coins, so their samples differ. Figures are bounded by the window we hold, not all-time claims.
How much to trust it
Seasonality is a real pattern in this data and a weak signal to trade on β both at once. Fifteen years is only fifteen observations per month, and a single outlier (Bitcoin's November 2013, say) can lift an average for a decade. That is why November shows a large mean but only a 60% win-rate: a few enormous months, not a reliable monthly gift. The honest use is calibration, not prediction β expect the mid-year to be quieter and the fourth quarter livelier, and size your expectations accordingly rather than betting on the calendar.
Methodology
- Data: our own daily closing prices, through 2026-09-01. Bitcoin runs the longest (15 years); the majors pool is the ten largest coins, each as far back as we reliably hold it.
- Monthly return: the change between consecutive month-end closes. A partial first or last calendar month is dropped, so no return is measured from a mid-month baseline.
- Average vs positive-rate: the average is the mean of those monthly returns; the positive-rate is the share that were green. The mean is skewed by early outsized years β the positive-rate is the more robust read.
Cite or share this study
The figures update as new data comes in. A link back keeps the citation live and lets your readers see the current numbers.
TheCryptoTools (2026). Crypto Seasonality: The Best and Worst Months. Retrieved from https://thecryptotools.com/research/crypto-seasonality/
Want to replay a specific run? The investment calculator plays any coin back day by day, and the drawdown study shows how far these same assets fall from their peaks.
Frequently asked questions
οΌWhat is the best month for Bitcoin historically?
Over 15 years of daily closes, November has been Bitcoin's strongest calendar month, averaging +36.5% month-over-month, with 60% of those months closing green. Averages are pulled up by crypto's early explosive years, so treat them as a tendency, not a forecast.
οΌIs September really bad for crypto?
It has been the softest patch for Bitcoin: September averaged β1.4% and closed green only 43% of years. Across the 10 largest coins the weakest month was June (β8.6%). The mid-year lull is real in the data β but it is a tendency across a small number of years, not a rule.
οΌIs 'Uptober' a real thing?
In the data, yes-ish: October averaged +14.4% for Bitcoin with 67% of years green, and the fourth quarter is historically its strongest stretch. As always, a handful of large years drive the average, so it is a pattern worth knowing rather than a trade to bet the farm on.
οΌCan I trade crypto seasonality?
Cautiously, if at all. Seasonality is a weak, backward-looking signal built on relatively few years, and one outlier month can move a whole average. It is more useful for setting expectations β knowing the mid-year is often quiet and Q4 often lively β than as a standalone trading strategy. This is general information, not financial advice.