// research
Uptober: is October really bitcoin's best month?
On Binance spot, bitcoin rose in 7 of 9 Octobers from 2017 to 2025, with a median October of +11.00% against +0.24% for the other months. Nine Octobers are a tiny sample, though: shuffle the same 109 monthly returns at random and at least one of the 12 months looks this good 60% of the time — and ETH shows no October effect at all.
6 min read
"Uptober" is the crypto meme that October is bitcoin's month. On the data this platform has — Binance spot since August 2017 — October is indeed at the top of the table: the best average month and the second-best median. What the data cannot do is tell that apart from luck, and the check on ETH points the other way.
Data and method
- Prices: Binance spot BTCUSDT and ETHUSDT daily candles, 2017-08-17 → 2026-10-19 — every daily candle the platform has for these pairs, with no missing days.
- Monthly return: the last daily close of a month against the last daily close of the month before. That gives 109 complete months, September 2017 to September 2026: nine of every calendar month (ten Septembers), and nine Octobers, 2017–2025.
- October 2026 is not finished at the time of writing: from the September 30 close to the October 19 close BTC was at -2.02% and ETH at -10.35%. It is in none of the statistics.
- Backtests: runs cannot start before 2020, so the six Octobers 2020–2025 are also run as buy-and-hold backtests on the strategy "Lump sum: buy once, hold" (#35, the one from DCA or lump sum): 9,880 USDT of a 10,000 USDT account buys BTC at the October 1 open and sells at the October 31 close, with a 0.1% fee and 0.05% slippage on each side. A run therefore shows slightly less than the price move: fees and 120 USDT left in cash.
- Luck check: the statistics of the nine real Octobers against nine months drawn at random from all 109, and against a shuffle of the month labels, 100,000 and 20,000 draws.
Every month, ranked
Up = the number of years the month ended higher. Rank is among the 12 months, by median.
| Month | BTC up | BTC median | BTC mean | BTC rank | ETH up | ETH median | ETH mean | ETH rank |
|---|---|---|---|---|---|---|---|---|
| January | 5 / 9 | +0.70% | +3.91% | 5 | 5 / 9 | +0.06% | +15.43% | 8 |
| February | 6 / 9 | +0.41% | +6.91% | 6 | 6 / 9 | +8.12% | +4.00% | 2 |
| March | 6 / 9 | +5.45% | +2.77% | 4 | 6 / 9 | +7.17% | -3.27% | 5 |
| April | 6 / 9 | +11.81% | +10.25% | 1 | 6 / 9 | +7.22% | +17.62% | 4 |
| May | 4 / 9 | -3.50% | +1.38% | 8 | 5 / 9 | +0.19% | +9.75% | 7 |
| June | 3 / 9 | -5.93% | -5.26% | 10 | 2 / 9 | -8.62% | -11.58% | 11 |
| July | 7 / 9 | +8.04% | +9.69% | 3 | 5 / 9 | +11.22% | +16.50% | 1 |
| August | 3 / 9 | -6.49% | -1.48% | 11 | 4 / 9 | -7.39% | +1.63% | 10 |
| September | 4 / 10 | -4.31% | -2.08% | 9 | 4 / 10 | -9.06% | -6.85% | 12 |
| October | 7 / 9 | +11.00% | +18.11% | 2 | 6 / 9 | +0.74% | +5.90% | 6 |
| November | 4 / 9 | -7.10% | +5.13% | 12 | 5 / 9 | +8.00% | +7.58% | 3 |
| December | 3 / 9 | -3.00% | +6.33% | 7 | 4 / 9 | -0.66% | +7.17% | 9 |
For bitcoin, October has the best mean, +18.11%, the second-best median after April, and shares the most up years, 7 of 9, with July. Across all 109 months BTC ended higher in 58, and the median month was +1.96%; the 100 months that are not October had a median of +0.24% and a mean of +3.36%.
ETH is the opposite picture: October is sixth by median and seventh by mean, and its median of +0.74% is the same as the +0.72% of ETH's other months. If October were special for crypto as a whole, the second-largest coin would show some of it.
Six Octobers as backtests
| October | BTC, price | BTC, backtest | ETH, price |
|---|---|---|---|
| 2017 | +47.61% | — | +0.18% |
| 2018 | -3.84% | — | -14.69% |
| 2019 | +10.27% | — | +0.74% |
| 2020 | +27.97% | +27.30% (#690) | +7.40% |
| 2021 | +39.88% | +39.05% (#691) | +42.88% |
| 2022 | +5.50% | +5.13% (#692) | +18.36% |
| 2023 | +28.47% | +27.79% (#693) | +8.60% |
| 2024 | +11.00% | +10.55% (#694) | -3.21% |
| 2025 | -3.89% | -4.14% (#695) | -7.17% |
Five of the six Octobers that a backtest can cover ended higher. Two of them did most of the work:
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2021-09-30 → 2021-10-31
- Return
- +39.05%
- Max drawdown
- 11.42%
- Sharpe
- 6.25
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2023-09-30 → 2023-10-31
- Return
- +27.79%
- Max drawdown
- 4.36%
- Sharpe
- 6.82
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2020-09-30 → 2020-10-31
- Return
- +27.30%
- Max drawdown
- 2.69%
- Sharpe
- 8.30
- Win rate
- 100.0%
- Trades
- 1
The small ones and the one loss — October 2025 also had the deepest dip inside the month, 14.48%:
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2024-09-30 → 2024-10-31
- Return
- +10.55%
- Max drawdown
- 4.83%
- Sharpe
- 3.29
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2022-09-30 → 2022-10-31
- Return
- +5.13%
- Max drawdown
- 6.31%
- Sharpe
- 2.05
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2025-09-30 → 2025-10-31
- Return
- -4.14%
- Max drawdown
- 14.48%
- Sharpe
- -0.97
- Win rate
- 0.0%
- Trades
- 1
In two of the best BTC Octobers, ETH bought the same way made far less:
Backtest
Lump sum: buy once, holdETHUSDT · 1d · Spot · 2020-09-30 → 2020-10-31
- Return
- +7.00%
- Max drawdown
- 7.54%
- Sharpe
- 1.90
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdETHUSDT · 1d · Spot · 2023-09-30 → 2023-10-31
- Return
- +8.19%
- Max drawdown
- 11.09%
- Sharpe
- 2.68
- Win rate
- 100.0%
- Trades
- 1
How much of it could be luck?
Nine Octobers is nine data points. The question is how often nine months picked at random would look as good.
- Drawing 9 of the 109 months at random, the median came out at +11.00% or more in 14.3% of draws, and 7 or more of the 9 were up in 11.5%; both at once in 6.3%. The mean of +18.11% is rarer — 2.2% of draws — but it leans on one month: October 2017, +47.61%, in the middle of that year's run-up.
- But October was not chosen in advance — it was picked because it looks best. Shuffle the month labels over the same 109 returns, and in 91.7% of shuffles at least one of the 12 months has a median of +11.00% or more; in 60.0% one of them has that median and at least 7 up years too. With twelve months to choose from, one of them looking like "Uptober" is what chance produces more often than not.
- Without 2017, BTC's October stays near the top: 6 of 8 up, median +10.6%, mean +14.42% — first by mean, second by median after April again.
Conclusion
- On Binance spot 2017–2025, October was bitcoin's best month by mean (+18.11%) and second by median (+11.00%), up in 7 of 9 years; five of the six backtested Octobers 2020–2025 made money.
- That is weak evidence: picking the best of 12 months from nine years of data finds a month this good 60% of the time by chance alone.
- ETH, the obvious check, has no October effect — its October median is the same as its other months.
How to read the run cards: backtest metrics explained. Another calendar question on the same data: the best day of the week to buy bitcoin.
Test your own idea
Create a free account, start from a template or from your own rules, and read your first report.
More articles
- Does high funding predict crypto crashes? 8 coins, 2020–2026BIS economists found that high crypto carry predicts crashes. On 8 Binance perpetuals, 2020–2026, crashes came more often — but after bigger gains.
- Crypto momentum after the paper: a weekly test on 8 coinsLiu and Tsyvinski found weekly momentum in bitcoin up to 2018. On 8 coins in 2020–2026 the one-week signal lost to holding; four weeks cut drawdowns.
- Does price move the hour before funding? 8 coins, 2020–2026Do traders sell before paying funding? The hour before each settlement on 8 Binance perpetuals, 2020–2026: no reliable move, and fees took what was left.
- Bitcoin after the halving: 2020 and 2024 backtestedDid buying BTC at the halving pay? A year after 2020: +554.04%, top 2.5% of start days. After 2024: +30.45%, below the median. Two cases, no pattern.
- Weekends and time of day: when does bitcoin move?Do bitcoin gains come on weekends or outside US trading hours? Hourly BTC and ETH data since 2020, and session backtests: with fees, -99.87%.